Implementing Limits on Grant Support to Strengthen the Biomedical Research Workforce

Posted

NIH realizes that, as stewards of the American investment in biomedical sciences, we must do all we can to protect the future of the biomedical research enterprise, taking additional measures regardless of our budget situation. In the opening pages of this blog, we noted that our increasingly hypercompetitive system is threatening the future of biomedical research and of the hundreds of thousands of scientists who we look to for discovering tomorrow’s cures. This is a strange irony, given that the last 25-50 years have been times of extraordinary discovery and progress in basic, translational, and applied science. Death rates from cardiovascular disease have plummeted, and death rates from cancer are falling steadily. Scientists have a much deeper understanding of human biology to the point where this knowledge can drive the design of drugs and biologics. Big data and high-throughput technologies now enable rapid development and testing of hypotheses that previously would have taken years. The successes are myriad. But so are the problems, problems so real that some have gone so far as to write, “It is time to confront the dangers at hand and rethink some fundamental features of the US biomedical research system.”

In these pages and elsewhere we have seen:

  • Concerns that the core problems besetting biomedical research are “too many researchers vying for too few dollars, [and] too many postdocs competing for too few faculty positions”
  • Data showing that since the NIH doubling ended, the number of scientists seeking NIH funding has increased at a much higher rate than the number of scientists NIH funds
  • Concerns that we are not paying enough attention to the number of investigators we support, as, given the unpredictable nature of science, we are more likely to generate transformational discoveries by funding more laboratories and research groups
  • Data showing that the supply of scientists continues to outstrip demand
  • Data showing that a relatively small proportion of scientists are receiving a large proportion of available funds; other data showing that there may be an increasing concentration of funds going to relatively few institutions
  • Data and models showing that the scientific workforce is aging at a much more rapid rate than the general workforce, leading to concerns that promising younger and mid-career investigators risk being crowded out
  • Data suggesting that scientific productivity tracks only weakly with funding; larger scientific groups or greater degrees of funding may not generate as much additional scientific output as expected due to the impact of diminishing returns
  • Data and concerns suggesting that the group of scientists who are most seriously affected by all these trends are young faculty, who see the need for funding as the biggest threat to their long-term success

These many concerns have drawn much attention – and perhaps of greatest concern is the wellbeing of the next generation of scientists. The US Congress, in the 21st Century Cures Act, has called on NIH to develop and promote policies that will attract and sustain support for diverse groups of outstanding young and new investigators.

What can be done to relieve the pressures of hypercompetition, to offer a brighter future for today’s early and mid-career investigators, and to minimize the impact of diminishing returns? How can we increase the number of independent early career scientists and stabilize the career trajectories of those who do high quality work? While we have seen some success with the implementation of our Early Stage Investigator policy, the concerns persist.

Going back to the opening pages of this blog, we noted a few of the recommendations that others have put forth: these included capping support for individual laboratories, capping salary support, supporting programs as opposed to projects, supporting more staff scientists, raising post-doc salaries, training scientists for non-academic careers, and assuring more efficient funding of expensive core facilities.

Shortly after we released that post, FASEB, which represents 30 scientific societies that count over 125,000 members, released a detailed report with extensive recommendations on how to sustain discovery in a strained biomedical research system. These recommendations included a call on research sponsors to “monitor the amount of funding going to a single individual or research group to ensure a broader distribution of research funding.” The report goes on to say, “Limiting the amount of funding awarded to any individual scientist or laboratory would enable more people to be actively engaged in research. With more ‘hands at the bench,’ the number of ideas would increase, and this could expedite progress in many areas of science.” A possible cap could be $1 million of RPG funding – such a cap could potentially free up enough NIH money to fund an additional 2,000 scientists. Kimble et al wrote, though, that any “redistribution” of funds from well-funded to unfunded (or nearly unfunded) investigators “will be painful, especially for established senior investigators, but necessary to support the next generation and cutting edge research.”

We and others have noted, however, that focusing on money alone as a measure of support may be problematic, as different areas of research entail different levels of expense: it is inherently more expensive to run clinical trial networks and large-animal research facilities. Therefore, we developed and described a different measure of grant support, which we called the “Research Commitment Index,” and will now refer to as the “Grant Support Index.” This is effectively a modified grant count, one that allows for different cost scales across diverse types of research, while at the same time accounting for the differing levels of intellectual and leadership commitment entailed by various NIH grant mechanisms. Consistent with what others have found with money or with personnel, we find that increasing levels of the Grant Support Index are associated with diminishing incremental returns.

Again — What can be done to relieve the pressures of hypercompetition, to offer a brighter future for today’s younger and mid-career investigators, and to minimize the impact of diminishing returns? How can we increase the number of early career funded scientists and stabilize the career trajectories for those who do high quality work?

NIH Director Francis Collins has announced that NIH is proposing several steps to set us on a stronger, more stable path, and to assure that NIH is maximizing the impact of the public dollars we spend. 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.

In our conversations over the next weeks and months, we will need to consider carefully a number of issues and details. How should the Grant Support Index be calibrated? Should we assign more or fewer points to certain grant mechanisms? What headline metrics should we follow? How will we know whether we achieving desired effects of funding more early career investigators and stabilizing their trajectories? How do we assure that we don’t inflict unintended harms on scientific progress, on the productivity of highly productive consortia, or on the stability of the research ecosystem? When would it be appropriate to allow exceptions to caps on individual researchers? How will decisions on exceptions be made? And what analogous steps should be taken with NIH’s Intramural Research Program?

We recognize the serious challenges and dangers that we face given increasing degrees of hypercompetition and scientific complexity. Yet, NIH is committed to assuring the robustness and stability of the next generation of investigators. We have a responsibility to the public to assure that we are optimizing the use of our limited resources to obtain the maximum impact possible. We look forward to working with you to figure how best we can use the tools available to us to “bend the curves,” including resetting expectation on support provided to any one research group. Finally, we will monitor and track data to assess progress, make corrections as needed, and mitigate unintended consequences.

418 Comments

  1. This is a plausible initiative…. productivity and scientific impact may be the best way to select grant recipients rather than concentrate founding resources in “big labs” with modest scientific outcomes. A better distribution of $$$ may speed up new discoveries and enhance collaboration btw emerging labs and “big labs”

    1. NIH should focus on how to better review the scientific merit of a grant, rather then thinking how to distribute money like a communist!

      Is NIH saying that a outstanding scientist should limit to have only 3 great idea?

      1. As it is now the ‘wealthy’ labs dictate what the NIH’s priorities are as they sit on various committees that advice the NIH as to where to put there money. In general, this has led to a ‘club’ type atmosphere were individuals within the club support one another. The labs received ridiculous amounts of money and published few and poor quality papers in return for more than 30 million in funding. Where’d the money go? What did they achieve? Hmmm. Personally, I worked in 2 so-called great labs (e.g. well funded, ~20 researchers); the PIs were average at best scientifically but surrounded themselves with younger, aspiring researchers with novel ideas that they happily took for themselves in return for empty promises.
        The cap is a great idea! They should also audit projects so that PIs have to show what they discovered for X amount of dollars. Personally, I like the idea the DoD uses that requires grant writers to be unknown to reviewers – let the best ideas get funded! Anything that can be done to improve mentorship qualities would also be appreciated.

        1. @Progress I totally agree. As others here have pointed out, 2 R01 are plenty of money for great research. Hoarders need to retire.

          1. 2 R01 – that really depends. If you do animal imaging and part of your grant is involving one or two clinical aims with a clinical trial those R01s don’t cover that much in terms as having a”big lab”, maybe own salary, postdoc (partially) and tech (partially), the rest goes into anime costs, imaging costs and costs for the clinical study!

    2. This article’s author seems to assume that the senior scientists or outstanding scientists in this country are not struggling for funding. Is this author living on earth?

      1. I don’t think this is what the author is saying at all. His point is that some lab receive way to much funding while there is little left to go around. I believe he is actually supporting your view that outstanding scientists are being left behind.

    3. NIH should focus on how to improve the review of grants. It is a communist idea to try to control distribution of money!

      A good science is a good science. It should not be limited by the scientist age, a number of grants, or other scientist is closing his lab.

      Current limit on federal funding is a good thing, because it will eliminate those labs that do not have good ideas.

      1. Easy… prohibit any reference that allows to identify who the PI on the grant is. Good ideas should stand on their own two feet. Then administratively review biosketches to check whether the PI does not have serious productivity issues before issuing the NOA.

    4. It always amazes me how easily people begin shouting “communist”. The main argument behind this cap is that data shows that increasing the funding behind a certain point produces diminishing returns. I agree that the best scientists may have more than 3 good ideas, however, the data seem to indicate that tackling more than 3 of them at once is inefficient. If that is indeed what the data show, we should pursue action that reflects that, just like we do when trying to solve a question in science.

      1. Excellent point Johan. A related point is that a lab chief can not effectively direct or mentor a huge number of people. If you put 30 great postdocs in a lab, some of them will produce great work, but it is unlikely that they are being effectively mentored by the lab chief. They would also produce great work (and the numbers might argue greater work) in smaller groups or running their own labs. Capping funding to groups allows the possibility that more funds can go to smaller labs and early investigators and the data seem to suggest that they will provide better value for the NIH dollar than the 4th or 5th R01 to a megalab.

        1. Look at the superstars at NBA. Similarly, we should also appreciate superstars in science!! Being a scientist, you do need talent and expertise….

      2. I looked at the data the NIH are relying on re productivity decreasing with more money to lay a foundation for this initiative and they are not convincing or robust-the data would not meet NIH’s current initiative for reproducibility and robustness, I think.

    5. What makes you think the big labs have modest outcomes? Rather than having a hard cap on funds that doesn’t make sense at all, why can’t we limit funds based on research outcomes. My lab is big and we routinely produce high-impact papers at least a couple times a year in addition to nearly a dozen mid-level papers. I don’t get why people think big labs are unproductive. If you find an unproductive lab of any size, the funding should reflect that. I’ve been in a few different labs in my time. The big ones are big for very good reasons. They are good at what they do and they manage their labs well.

    6. It would be inappropriate to implement such a policy change without evidence of effectiveness. As a leader in science, the NIH should make only evidence-based policy changes. The risk is that the intervention may reduce the quality and impact of the science it funds. At the very least, a pilot study should be done. An even better alternative is a cluster randomized trial.

    7. While I firmly agree with supporting our young investigators I am strongly opposed to this GSI initiative which I believe is shortsighted and will backfire in the long run causing more young people to leave science. Consider for a moment that as a senior investigator I use my grant funding to support the work of young trainees on their F and K grants– support that I can’t continue if I am penalized for my grant portfolio. Consider also that I often bring young PI”s in my Center up through the ranks by inviting them to be co-PI’s on multi-PI grants during sensitive transition times when they wouldn’t typically be ready yet for stand alone R01s. I won’t be writing those grants anymore under this system. All my grants include considerable funding for Open source sharing of my reagents, cells, and datasets. These R01 grants help me share these essential tools with countless young investigators nationally that then can compile preliminary data to compete for their own R01s. Often I share these cells and tools prior to publication in an effort to get them out the door to the young research community. All that will end with this misguided GSI initiative as I’ll be penalized for my existing portfolio, will need to scale back and the first thing to be scaled back will be this essential support for sharing with young investigators. Please consider that we need to help our young trainees through many means, including funding productive labs that train these individuals. Punishing the success of well funded investigators only adds to the negativity and competitiveness of science, it only will spread the wealth very short term (next year) and will backfire within 5 years to worsen the attrition of investigators from academic medicine and science. We must continue to support our young and also reward success!

      1. I agree with this assessment of the proposed GSI. Senior investigators are an essential part of good science. We should not institute policies that devalue success. The best way to help young investigators is to provide them with the tools to do good research with senior investigators leading the way until the young investigator’s productivity has been proven. At that point, they should have no difficulty obtaining funds of their own. I hope NIH will convene a panel of senior and young investigators to explore the GSI more fully before implementing.

  2. With one child in his first year of college and a second just 2 years away – I can vouch that ‘STEM’ recommendations are still at the top of every high school career counseling office and at the top for suggestions for fields for incoming college students. Perhaps there needs to be a more targeted suggestions as the S in STEM should really no longer encompass the Biomedical research field as there is clearly not a need for this area. Physics and chemistry – yes. Biomedical research – not so much. We are doing a disservice to young people with the ‘STEM’ message. For example: https://www.ed.gov/stem This Department of Education website makes it appear that biomedical science jobs will be plentiful.

    1. I agree. The STEM initiatives should not emphasize the soft sciences, but rather math, chemistry, physics, and engineering. I would venture that the STEM initiative has had a similar effect as online universities – leaving unaware students jobless and in debt – probably especially students from underprivileged populations.

  3. Well, ignoring the arguments about diminishing returns, funding young investigators and negative impacts on some of our individual research programs, efforts to support and regrow the “middle class” of NIH funded investigators are long overdue. Reflecting the last several years when most institute paylines have been well below 20th percentile, many formerly productive, single R01 labs have lost their key funding and are now dead in the water. Most of us would willingly take a hit to our research programs if it meant more of our colleagues could keep sailing. How this gets done is another matter.

    Whether limiting PIs to a specific number of “points” is the best way to achieve this is hard to say. I am not sure why a typical handicapping approach cannot be used where getting a third, fourth or fifth R01 is successively harder, like via a sliding scale with respect to payline, so that colleagues can feel the “best science” is still getting funded?

    If the payline for a first/single R01 PI was no worse than 20th percentile but it took a 2nd percentile (likely the best grant reviewed by the panel) to get a PI’s fifth R01, I don’t think there would be as much concern about diminishing returns. In turn, proposals for a fourth or fifth R01 would only be submitted when a PI really had something outstanding to propose.

    1. This is a great idea and a balanced approach to the trade-off between “spreading the wealth around” and not forfeiting game-changing ideas because a particular PI is at the top of his/her field. Lets also not forget the other implications of reducing funding to super labs on the same early career investigators we are trying to protect. Many new investigators previsouly sought out post-doctoral training in these prestigious labs due to the invaluable training that could be received there. There will be less of these positions available if we intentionally reduce the supply of funds. If the ultimate goal of these policies is to foster the future workforce, lets do things to ensure that we are sending money to labs that have a strong track record of producing future independent successful scientists.

        1. There have been a number of comments to the effect that the data showing diminishing returns at higher funding levels are not sufficiently robust. However, there are absolutely no data suggesting the opposite: that increasing funding levels result in increasing productivity, and particularly not with a slope anywhere close to unity.

          Everyone who has served on study section knows that (a) there is not a reliable difference between say 10th and 20th %ile, and (b) that name recognition swamps the measurement noise in that critical 10-20%ile range.

          The onus should absolutely be on those who would defend the status quo to marshal statistical evidence — any statistical evidence at all — in support of their position. So far, I’ve not seen anyone even try to do so. The best that these researchers (to a one, funded by taxpayer dollars, of course) can do is to screech about “socialism.”

    2. seems like a good idea to me. My impression was that the original idea of looking much more closely at PIs with $1M/year DC was essentially going to be that. It would be good to know what happened when they tried to implement the policy that created the need for something like this index which has serious issues in my opinion.

    3. I think this is a very balanced and thoughtful response. I also feel that his proposal of only funding 3 or more NIH grants only if they get really outstanding scores (e.g. <5%) is a much better idea than automatically capping grants at 3. Scientific competence and quality like all professions is a bell shaped curve and we need to enable the truly exceptional scientists to really thrive while maintaining sufficient funding to support a wide breadth of excellent scientists and especially rising young stars.

      NIH also needs to improve the quality of peer review as their efforts to involve far more individuals in the review process including excluding outstanding scientists once they have met a lifelong quota, has resulted in numerous mediocre scientists serving as charter members of standing review panels. I agree with promoting diversity on review panels but not at the expense of review quality.

    4. This is a REALLY interesting idea that seems to retain the spirit of the intent of the GSI while getting around some of the common critiques. At first blush, I like it a lot!

      Guessing there are a bunch of math/modeling experts around here somewhere. Can anyone put together plausible mathematical models to help predict how various strategies might play out over time, as particular factors vary in one direction or another?

  4. Q: are we possibly confusing “elite” and “old” – there should be data on this readily available and something we can model easily. If the heavily awarded PIs, the ones in the 6% we are targeting with the GSI, are all old then we have a problem but if they aren’t then we would be yanking support for the best scientists America has. For example, success rate may not be the same as the GSI capped award level (you could have lots of successful old-timey double-R01-ers but then could come to find out that the “6%ers” – the quadruple R01-ers are skewed to the hotshot youngsters that we desperately need, the McArthur Geniuses and such). An alternative explanation for what appears to be “elitism” could be the fact that with more % of the population going to college and graduate school that there has been a broadening of the base of schools (age could still be a factor but the 6% cap targeted by the GSI may not affect that if it age is a smaller component)) that are now trying to compete with the Stanfords and Harvards (i.e. think of all the 2y and 4 y community colleges that have grown-up into massively enrolled universities with graduate programs in just the last decade or two).

    I hope we pilot the first attempt in a limited way as I envision a lot of ping-ponging as we overweight one grant mechanism for capping and drive PIs into other mechanisms then have to adjust the coefficient for capping to drive them back etc.

    1. There are some good ideas here including raising the threshold for funding after the first RO-1. I also think that extending the duration of grants would reduce the number of redundant applications, lower the workload for study sections which could focus on progress rather than promise. Most important, the NIH funds way too many mediocre scientists, some of whom do more harm than good with their work. Until NIH figures out how to incentivize the most qualified reviewers, quality control of their product will continue to deteriorate. The other problem, much bigger, is that the institutes by and large have not done a good job of identifying the problem they want to solve. There is a tremendous amount of very good science being funded that has nothing to do with the reason money was given to NIH. Better focus and quality control of research will lead to better health outcomes for the taxpayers.

      1. Many research institutions and universities have fairly high expenses in terms of indirect cost rates and relatively high salaries for full professors. The default indirect rate for those without negotiated agreements is 40%. If this were applied across the board, many more proposals could be funded. As for payscale, years ago, university professors worked for salaries that were considerably lower than corporate scientist salaries in exchange for academic freedom to work on projects of interest. Now, the university salaries have caught up or exceeded industry salaries. Perhaps it is time to reduce the maximum reimbursable salary instead of continuing to increase it.

      2. The threshold to funding after the first R01 is already raised – by decreasing it for young investigators. A communist approach to science is not a good way to conduct science – deny funding to a great application because the PI has more than the allowed points and rather give it to a less good proposal just because that person doesn’t have all the points yet? And we have the percent effort to take care of that in a way.

        1. As an example, let’s assume that there’s only enough money to fund 8 proposals in this round. Are you saying that you or other reviewers on study section can confidently distinguish the difference in quality between the 8th best proposal versus the 10th best proposal and therefore the 8th best should be funded over the 10th best even though the former PI already has 3 R01s whereas the latter PI has 0-1 R01s?

          1. This is a very good point and a reminder to consider the true context in which the “rules” will be applied. The reality is that there are a lot of good ideas, a variety of angles, etc. and it can be hard to judge, based on the presentation in a grant application, where the real breakthroughs will come from.

    2. A elitist system has no reason to exist, no matter if it is about Harvard or MacArthur recipients. In fact I worked as a grad student in the lab of a MacArthur “genius” who also had a Howard Hughes grant. He was a self-centered individual from whom I did not learn much. His papers were not ground-breaking, and after 7 years he lost the HH grant. Now he has 2 people in his lab, if even that, a huge, deserted lab. So let’s not bank on these “geniuses”, young or old. If this guy was at his peak now, he would probably hoard grants just based on his name, and prevent others from getting anything. If anyone has more than 5 people in the lab, it is unrealistic to think that they follow everything that is going on. If they have more ideas than resources, they should collaborate with other independent researchers willing to pool resources, not hoard everything for themselves. I am sure that if something is worth pursuing, others will jump on board. A lot of researchers have their Evrika moment, when they find something meaningful, after which they sink in mediocrity. This usually happens because most are interested in the financial aspect of their position rather than having a genuine passion for science. Young ambitious researchers later become a drag. These guys must not keep getting grants just based on their name or university where they work. Unfortunately, this happens a lot, and when someone makes a name, they keep having an undeserved lifetime influence both on grants and on paper reviews, actively preventing others from challenging them. This is the death of science.

    3. The problem is the “diminishing returns”. Elite or not, more grant money does not equal more research productivity. If the productivity of that 4th R01 will be equal or better than the first one, sure, go ahead and fund that lab even more. But this is an extremely rare situation, and the productivity/researcher ratio of a lab is generally significantly lower for a 30 researcher lab vs. a 10 researcher lab.

    4. Why do you assume that “old” should be a factor in anything? There are plenty of old researchers who are highly productive and innovative and plenty of mid-career researchers who are turning out mundane research. There may indeed be a problem of some sort of “halo effect” that makes it easier for “elite” labs to get funding for incremental research, but the way to address this is through the review mechanism.

      I do agree that there is a great danger of “yanking support for the best scientists America has”. So we need to be careful in how any changes are made.

      1. Frankly, I have seen the quality of grants is going down in recent years. The key for NIH is how to make unfunded investigators become more scientific competitive and generate better scientific ideas. In addition, improving our current review system is another important task.

        1. Frankly, I have seen that the quality of the grants that get funded has decreased, while those that should be funded are not (as a reviewer due to extreme and unchecked biases).

          1. then you need to speak up and we should work on how to improve the review system. it is just wasting time to make nonsense arguments… particularly for a meaningful “reviewer”

          2. I thought that I just did! And I do so during reviews, as sometimes also do the SROs, but it doesn’t seem to matter much. I agree that the review system needs to be revamped, kind of my “nonsense” point.

          3. That is the “nonsense” point, since you did not speak to the key persons at NIH! Sending the letter to the head of CSR, being meaningful and effective. Don’t waste your time on Blog here.

  5. Follow-up: Any GSI we come up with is still a poor derivative of what we really care about which is awarded grant dollars and percent time of PI’s on grants. If a GSI is intended to address old PI’s consuming too much of the pool of grant monies AND we can’t target them directly for fear of age discrimination then the answer is not new math and new coefficients but to just adjust the payline (or definition – like how many awards they can get and retain status) for new PI’s/ESIs. But surely this was the first thing anyone thought of so what is wrong with that approach?

  6. I think the idea is a good step forward – another idea worth considering is creating a new class of grants that will only be accessible to investigators who have no current NIH funding (or only 1 RO1 level grant). Such grants could keep labs from dying allowing important continuity – among many other benefits!

    1. Brilliant idea! I am about a year away from having a dead lab (it was small in the first place) and cannot get close enough on my R01s. I am spending all my time and money trying to get more and more preliminary data to satisfy grant reviewers, which means I cannot focus and am looking at more than 15 years of independent funding going down the drain. I have seen large labs get 5-year R01s that never resulted in a single paper, yet their funding has continued to increase. I feel like, having a smaller lab, I can focus more, do a better job of mentoring my students (instead of just dumping them on a postdoc), and generate a higher ROI (since I have to be a good steward of limited funds). Frankly, I (and many other mid-stage PIs like myself) need help. Otherwise, when the senior PIs do finally retire we will all be out of science and there will be no-one left to do the grunt work of science (editorial boards, grant review, etc.) or with the expertise to mentor.

    2. I agree strongly with this idea “…creating a new class of grants that will only be accessible to investigators who have no current NIH funding (or only 1 RO1 level grant). Such grants could keep labs from dying allowing important continuity..” This would effectively improve pay lines for young investigators/those with a single R01, without penalizing more senior investigators who can have highly productive 3-5 RPGs based up on some metric that may reflect reality for some, but not all, more senior investigators. As someone who helps new & mid-career investigators learn how to write grants that make the pay lines, I also see reviews that are clearly from very senior investigators who sit on review panels that indicate they are not open to new, innovative ideas that more junior investigators/those with smaller labs present in their grants – essentially funneling funds to their more senior peers (whether that is their explicit intent or not) whose ideas they are more familiar with, and whose writing styles are more familiar to researchers “of a certain age”.

    3. Agreed. There is a need to maintain the established investigators who are not part of the so-called elite. Labs are dying out, floors are empty. The elephants in the room include funding early career researchers who do not progress, “mafias” and the “sparkly syndrome” (reference to “The Secret of NIMH”). Finally, an new approach to review that allows for immediate Q&A between reviewers and applicants might be an appropriate way to ensure that the review process is fair (I am a reviewer).

    4. Ken and Ian, you are both right. Not only the number of grants, but also the amount of money per grant should decrease. There is no valid reason why some researchers, especially in “brand name” universities, are allowed to hoard huge resources. Funding should go directly to those who do the work, not to an “overseer” who makes 200K+ salary. If universities want to award huge salaries, that is their problem, they should do that from their own funds, not from public funding. Awarding large salaries stimulates the greed of some researchers to hoard grants to meet the ceiling of their salary. Many research hospitals offer salaries of 300K+, but much of that comes from grant funding. This is ridiculous. The hospital should pay the entire salary for their medical services, not expect MDs to get grants to reach their maximum income. If an MD is passionate about research, they will apply for grants even if that does not bring them additional income. If they apply for grants only to supplement their income, then that is a problem, because it sheds doubt on the quality of the research.

      1. There is an NIH maximum salary of $185,000 per year. While some institutions certainly pay investigators more than this, no more than this amount can be recovered by NIH salary recovery.

        1. right, but if the hospital technically offers a salary of $300k, but only pays 100k, leaving it up to the MD to make up the difference from grants, that is a problem, because it forces an artificial competition for grants.

          1. That makes no sense at all. If the researcher is as productive as their salary would suggest (i.e. 300K/yr), then shouldn’t they get paid more than a researcher that is less productive? Artificial redistribution of resources has rarely, probably never, resulted in better science.

          2. Not sure why singling out MD to penalize the high salary. With my NIH funding (I am MD), fundings cover about 50% of my salary, and I make the other 50% through clinical service (seeing patients clinic and in hospital) while maintaining two R01s (one large clinical translation and the other bench projects). If a PhD is asked to teach to cover 50% of his/her salary in addition to 50% funded salary to cover the whole salary, I would say there will be some incredible push back. The funding issue is delicate. Conducting a competent review to fund “good” and “competitive” science while eliminating not so good science (production inclusive in this term) has always been hard. We had had one project scoring better than 5th percentile while another project getting triaged in a same review cycle… My point that review process may sometimes seem unreliable and random, but it does produce competitive environment. I would say top 20% funding seems good middle of the road comparing late Bill Clinton years and past 10 years… I am still interested in hearing details on how NIH will balance these issues.

  7. The problem is how to measure productivity related to dollar amount. Is it the number of publications or scientific advancement in the true sense? It seems that most of the NIH grants are clustered at the East and West costs. This is obviously because of big name schools with big name investigators and plentiful of resources.

    Since NIH funding is greatly dependent on peer review merit scoring of R-grants, increasing the younger workforce in this business will require a system of funding senior investigators paired with a young investigators with more funding dedicated toward the young investigators. This way young investigators will be mentored and nurtured by senior investigators to be more competitive.

    There should be a vigorous national and regional dialogue to address these issues.

    1. I agree. I’d love to see the Ks done away with and R01s for new PIs require a 10% senior investigator mentor who can guide them on things like handling a budget, hiring staff, overseeing staff and time management; along with actual scientific mentoring. It would seem to me that it would offer a graceful and dignified transition for senior investigators nearing the end of their careers. And it would serve new investigators better than 5 years of being mentored on a project that they can’t afford to do on a K budget.

        1. Agreed. The KOI is not a good mechanism – way too small an amount of money, unrealistic training needs have to be listed. Mentored 3 or 4 year R01-type of grant with some training budgeted for would be much better. Also, should be allowable for a senior post-doc to apply (who would then get an assistant prof position if successful).

    2. I like this idea… More like continuous, more years of apprenticeship and mentorship.. beyond just 5 year Ks…

  8. From a public investment perspective, the diminishing returns on investment makes a solid argument to implement this policy. Social science research shows that diverse groups are better at solving problems. So it seems the best way to solve a biomedical research challenge is to spread the resources more widely and get more brains n the problem. The lack of diversity of the biomedical research workforce has been stubbornly resistant to efforts to correct. Younger scientists are both more creative and more diverse, these are known facts of social science and demographics. This policy will go a long way toward increasing diversity of the research workforce. I think there may be an added benefit of democratizing the spread of NIH research investment to institutions that have difficulty competing because they may have newer programs with younger, more diverse faculty that are faced with competing against institutions with disproportionate advantage through older, more established faculty that are viewed, perhaps incorrectly according to emerging data, as better risks for investment and return. These institutions are essentially systematically discriminated against. More research in more institutions provides more role models and mentors for underserved students interested in STEM careers. One can imagine a scenario wherein established institutions may support more young people to preserve their edge in overall funding, and new or emerging programs will be better positioned to compete.

  9. This is not an effective way to increase productivity. Grouping grant mechanisms together suggests that they result in equal productivity. I would argue R01s with their modest annual budget result in much better productivity (paper/dollar) than for example P01s with their enormous budgets. The intramural program is not discussed at all. Chief’s at NIH can have budgets of $8 million/year, get renewed every five years with ease and nobody is talking about cutting this. Compare that to four R01s, which gives you $1 million/year, each has to get renewed every five years by limbo dancing beneath a 12% payline, and is now proposed to being cut.

    1. I completely agree. And I would think that grants like R03s and R21s provide even more output per amount funded and these grants last only 2 years. Assigning 3-4 points for every R03 and only 10 points for a P or U grant seems unreasonable. Please consider reducing the number of points assigned to R03s, R01s, and R21 grants (and other grants like them).

      1. Completely agree. Why should an R21 be assigned a score equivalent to 83% of an R01 when it provides only 25-50% of the funding per year and 3 fewer years? This doesn’t make any sense.

    2. I came here to say much the same thing. As a “basic scientist”, a 5 year R01 (and a couple of small, intramural projects) is more than enough to keep me busy, along with administration, teaching and local and national committee contributions. However, a colleague performing clinical trials needs a lot more money. I have no problem with this reality, except that we as faculty are judged by how many grant dollars we bring in . It is supposed to be the triumvirate of funding, teaching and service that are taken into account when applying for reappointment or tenure, but in reality it is $$$$$$. Just because my research happens to be less expensive (but no less important and of contribution to the field), I will, on paper, be less impressive than someone with a clinical trial. If this change at the NIH level is to be successful, there must be a change in how Universities ACTUALLY assess criteria for success in their scientists.

    3. It depends on the nature of the P01. One weakness of the NIH funding system is that it is not well-configured to evaluate and fund transdisciplinary research. The study section memberships are largely drawn from individuals whose grants have been reviewed by that study section or one very close. They are almost all very good scientists, but they generally tend to get most excited about work with methodological similarities to their own. These days to get a fundable score, two or three reviewers have to be very excited about the proposal. With the number of excellent proposals that are submitted, there are always outstanding proposals that are well within the expertise of study section members and they naturally get the best scores. PO1 study sections are convened to review a particular application and can be constituted to have the breadth of expertise required to evaluate it properly. I am sure that there are bloated PO1 awards that are not an efficient use of NIH funds. P01 awards can be as small as three normal RO1 grants, however, and can be very good value for money because they can fund transdisciplinary work that is novel and difficult to get funded through the R01 mechanism.

    4. I couldn’t agree more. Why aren’t more rigorous reviews of the intramural NIH scientists’ budgets required? I have seen so much waste in Bethesda while the extramural NIH grantees are starving, using 25 year old equipment and watching every penny. Why doesn’t NIH do a comparison of the productivity in terms of numbers of publications vs. all monies spent on the research (including salaries, outside contracts for animals, etc.) for intramural scientists vs extramural scientists??? I think that this would be VERY illuminating as the intramural budget is a big portion of the NIH total budget and is rarely discussed.

  10. What you forget in this endeavor that these days we need to push out many grants to be successful. I didn’t get anywhere for 2 years with any grants and all of a sudden got three. Didn’t plan for it, just needed to get funding and was desperate, pushing out up to three R01 per cycle. Now, I will have these ending nearly at the same time. So I need to push out many again in 3 to 4 years. What happens if I get one in year 4 of my 5-y grants and then am told that this is not happening because I already am over the 21 point limit? Getting renewals is so tough these days that you cannot count on it.

    1. I’m in exact same boat. Went from rags to riches in desperate attempt to stay alive, but could end up back to rags in no time. I’m midcareer (age 44) so with this implemented I worry I won’t be able to submit as many grants to “stay alive” because I’ll be at the “limit.” I publish plenty so the statistic about lack of productivity does not apply to me. This would not seem to incentivize productivity among the very productive.

      1. Same here. I had a gap in funding as PI for several months but I now (unexpectedly) have three R01s (2 of which are MPI) and an R03, which puts me over the cap. I have a couple of other grants that are under review in June (these were already in the pipeline as part of my desperate attempt to bring in funds as PI last year). I am mid-career (age 42 years). I would be very disappointed to be overlooked for funding after having worked so hard to write these grants. Is there potential for the grants to still be funded but to transfer PI-ship over to a more junior investigator on the grant? Would a meritorious grant be skipped over for funding even if the science is judged to be outstanding? I think this is a disservice to science. Rather, the NIH should offer the opportunity for PI to another investigator from the same set of investigators and the former PI could take on a senior mentoring role as co-investigator. I hope the NIH will seriously consider doing this instead of skipping over well-scored applications.

        1. Ditto! I am also mid-career (44). I wrote for years with no success and then received two R01s and an R21 within a 15 month span. Both R01s were scored below the 5th percentile and the R21 was scored beneath the 10th percentile–by three different study sections. I feel like I am really just getting started on my independent research career (I had a K), and would be severely hampered by the GCI (I am two points away). My research is patient oriented with hard to reach populations so additional funding allows me to leverage the resources that I already have. There is no “hoarding” of resources in research with tough to reach populations by any stretch of the imagination.

          Some have argued that the next step is to simply pair with a strong junior investigator, and allow them to be PI on a grant that I’ve written. However, it is difficult to pair with the most outstanding junior investigators at my stage. I can attract great postdocs, but the Assistant Profs prefer to work with the super senior PI, one who has had multiple R01s over a 20-year period, rather than someone who received their first R01 1-2 years ago. So that would leave me with more limited junior mentee options, at a stage in my career when I want to grow. NIH has numerous programs for current ESIs, and the most senior PIs are protected. However, those of us at the mid-career stage require support as well. If we are weeded out, who will mentor the next generation of ESIs?

  11. It’s unclear if a GSI is calculated on active awards (can’t be a PI on > 3 active R01s) or lifetime (can’t have more than 3R01s)?
    Also, would a renewal be treated the same as a new award.

    Overall, it is an intriguing idea. Look forward to hearing more about this proposal

    1. Yes! I think this is a very important distinction that should be discussed. Also, if you are at a 100% soft-money institution or a position, then is it fair to limit your lifetime employment potential?? Is it even fair to limit your concurrent support?

      1. Thank you for mentioning soft money “labs.” I work in a largely soft money center where if this policy were implemented it could seriously hinder the ability of our PIs to stay employed and continue their excellent, highly productive programs of research.

        I also was intrigued by the proposal to assign almost as many points for multi-PI awards as traditional single-PI awards. This could disincentivize team science collaborations.

      2. @Kate Grant support is NOT a lifetime right. I have been working at a research center where my contract is renewed every year, pending on results and on whether there is money available or not. If you want to keep doing research, you have to keep being competitive. As in a lottery, past results are not (or SHOULD NOT) be a guarantee of future funding. If someone had a good idea 10 years ago, you cannot keep funding that individual hoping that someday, in another 10 years, there will be another good idea. I find seriously flawed a grant award system based on considerations such as the university where the research is done, whether the PI has lab members to support or not, whether the PI has tenure or not, and other things that have nothing to do with the quality of research. If anything, good results at average universities should be valued more (get extra points) because it is more difficult to reach a high level of performance in less competitive places.

  12. Are there plans to also implement caps of support for investigators in the Intramural program? Obviously the funding mechanism is very different, but the argument about larger labs and diminishing returns is still applicable.

  13. Be very careful with: 1) Incentives, and 2) conflict of interests.

    1) I come from a socialist country, and what I’ve seen that happens in the real world is that spreading the money, unless it’s done under severe restrictions, is an idea of disastrous consequences in the long run. The solution has to answer the simple question “who makes a better use of the money?” The answer will be different depending on the development stage of an investigator. For example, a new PI could be making a good use of the money with a One-shot break through investigation, while a bigger lab could be considered good if it’s a well oiled machine capable of producing research and researchers.
    The problem with cap sizes is the same of socialism: what incentives has an investigator to grow if it has a maximum size set by the government? What happens in those economies? Yes, you are intelligent people and already know the answer: they go to other countries that allows them to grow, OR they accept their fate and either stop trying to grow, or they move to other areas where their talent will be more appreciated. Finally the result is the catastrophe, a crippled system unable to generate movement and in need of a complete restructuring.
    The question was never “if” or “how much” they have to grow, is “who” and “how” are they growing, therefore the question goes deeper in the impact of each investigator more than the mistaken optics of diminishing returns on the amount of papers produced.

    2) be careful on WHO you listen to, many advisors will ultimately represent corporations interested in SOME of the most lucrative research (ers). I’m not against private investment, but the real world shows that corporate interest is not necessarily aligned with the long-term view that science(tists) must have.
    Just as an example refer to t. Cruzi: there is a treatment, but it’s not FDA approved, is a cheap treatment, and for that reason nobody will ever research with private funding… But there’s people getting sick below the 35°N parallel.

    You want to know if it’s a good plan? Answer the simple questions first: we are taking money from someone that does a good job, who can do it better? But what does it mean to do a better job? What is a job? A paper? Seeding new researchers? Will we leave untouched the (infamous) role of science magazines? Will we leave untouched the systemic lack of result review? Why can’t we set a system of public tenders for matters of public interest?

    1. I agree. We need to strongly consider the merit of the science being proposed and its impact. Socialization of scientific funding sounds like a horrible idea. The well established
      investigators may actually be in their position to get continual funding
      for a reason…perhaps because they write better grants and are more innovative based on their experience. The ideas spreading around here I think are rather dangerous
      and pernicious. I also haven’t seen much discussion on peer grant review and the process…this has also become a broken
      system. Lots of funny business when it comes to who gets funding, even when grant scores are excellent.

  14. The median (~50%) of support to an intramural NIH is $1.27 million, or approximately 5 modular R01 equivalents, whereas only 6% of extramural NIH PIs have 3 R01 equivalents or more. Will this policy be applied to intramural investigators, or will they be excluded? If intramural excluded, what is the justification?

    1. Yes, lets focus on the intramural NIH. I am shocked at how much money intramural researchers receive vs. extramural. Why isn’t this a huge focus in terms of fixing funding rates? Let’s limit funding of intramural NIH researchers to the equivalent of a maximum of 3 RO1s-with administrative cuts this will be about $650,000.

  15. I have been thinking about this issue for a long time after observing the many young researchers (assistant professors) who were not able to generate extramural funding and who consequently, they were not considered for promotion and tenure. One fundamental problem with the current grant review process is that the competition for funding is not set up in an equitable fashion. That is, we have newly minted assistant professors who might have one graduate student and one post doc in their labs ( if they are lucky) competing with established investigators running labs with a stable of post docs and technicians. Of course, the already well-funded full professors will be able to submit a grant that is rich with preliminary data and that will consequently score much better than that of the young investigator.

    My suggestion is that we have different tiers of grants/pools of money that are committed to applicants at different career levels. The size of the grants could also be commensurate with the career stage of the investigators. That is, there could be grants of e.g. $300,000 over 3 years for assistant professors; $500,000 over 4 years for associate professors and some higher amounts/years of funding for full professors. In this way, each applicant would be competing at his or her own current career level.

    In this context, I recall sitting on a review panel many years ago for some obscure funding mechanism that was dedicated to providing additional support to already funded investigators. The logic of this approach eludes me now as it did then. I recall that the best proposal came (not surprisingly) from the lab of a well-known scientist in the cancer field who, by my calculation, had more than 4 million dollars of funding in hand. I was quite certain that the additional monies would be unlikely to have any significant impact on his/her research productivity.

  16. Three points.

    This is not an effective way to alter productivity. What should be eliminated are large NIH sponsored programs, which are expensive and rarely have the intended impact. Most large programs serve those who are funded by it. NIH should be fostering R-01 science, All labs are not created as scientific equals. and peer review, although imperfect, is the only way to included merit.

    Technically, NIH in its wisdom fails to see the critical details when implementing such sweeping policies:

    Considering these issues is critical.

    (1) I am on soft money, Lawrence Berkeley National Laboratory–there is no salary and no DOE support. I completely support my lab by my own grants. Therefore, I need one grant to simply support my salary. Capping my grant support will have a serious negative impact on my research and may lead to site closure. What is spread wider at a university will not go far at LBNL.

    (2) The indirect cost rate must be considered. At LBNL, we pay 30% more per individual than at academic institutions.

    (3) We pay indirect rate of 89%. The caps must include full cost recovery , i.e., normalized for indirect rates.

    (4) Laboratories base on animal work have double the expense of labs using cells. Unless you want animal work to disappear, you must determine a way to equalize effort. One way would be to devise a monetary limit for direct costs and calculate animals costs separately.

    These broad changes, although sympathetic, are never really effective, and by spreading the thin pot thinner, fail to accomplish their intended goal.

    1. @Cynthia Actually I find peer review quite flawed, especially for individuals who are not part of a network of connections. A much better indicator is the number of citations. Also, having a lab does not generate a lifetime right to obtain funding. Labs get started and closed all the time. A researcher without a lab who wants to start one has as much right to funding as a researcher who already has a lab. I understand your problem, but when funding is so limited, extraneous considerations should be eliminated. As a post-doc, I lost my job twice because the PIs could not renew their grants, but I found other positions and the changes proved very helpful. So having people in your lab is not a factor in obtaining grants.

    2. Yes, I agree….eliminate all of these special, large programs funded by NIH! Go back to funding almost all research via the most transparent RO1 mechanism

    3. Your argument is, “Science costs more at LNBL.” A reasonable response is, “The government should get the most science it can for each dollar, so soft-money institutions with low salary support and high overhead are the least efficient ways to purchase science.”

  17. This has to be one of the most short-sighted policies I have ever seen:
    1. We already have a system in place to determine how much bandwidth an investigator has, and limiting him/her to that number of grants. It is called percent effort.
    2. No one writes more NIH grants because it is fun. They do it because they need the money to fund the work they want to do. If the work is exceptional, and makes it into the top 10% which is where the paylines now are, then it should be funded.
    3. This punishes some of the best investigators. 90% of the advances in any field are made by the top 10% of investigators. If the work is first rate, and the investigator has the bandwidth to do it, it should be funded. Limiting the productivity of top notch investigators who can make real advances to “spread the money around” makes no sense.

    Rather than figuring out ways to kill one another off, we need to work with Congress to recognize that the NIH budget has been cut, in real dollars, by over 20% in the last two decades. That is the real source of the problem. The solution is a plan for sustainable NIH health, such as increases of 1-2% above inflation per year going forward.

    1. Amen ! this is a great argument – just enforce the %effort as it should be. BUT fund the best science!

    2. Data indicate otherwise. Turns of that PIs with large amounts of funding are not necessarily “exceptional”. Or not any more exceptional than PIs with a single R01. The data clearly show that a PI with 4 R01s produces somewhat less than 4 PIs with one R01 each. Having a large lab and an extensive track record increases your chances of _getting funded_, but does not improve the _outcome_ of the research.

      1. But the data on this are weak. Have you seen the actual data from NIH on this point? The data would not survive the current robustness push by NIH for extramural research.

    3. “90% of the advances in any field are made by the top 10% of investigators”.
      Totally agree. But you need to fund 100% to get these top 10% successful investigators. You don’t know whether one of the 90% PIs now will get into the top 10% later.

    4. the point is, the same 10% of top investigators can do an almost equally good job in productivity with half the grant dollars they receive from NIH.

      in other words, diminishing returns.

      1. you are not right – productivity cannot be simply measured by publications – the training large labs provide is never taken into account – this will just breed mediocrity – if that is what you want then fine –

        1. Training a lot of scientists without them having publications is another problem, clearly mentioned in the original article. Too much supply, too little demand.

          Besides, I don’t know a better measure than publications to judge productivity. It is still the best measure we have.

          The data is clear, extremely large labs are less efficient and less productive.

          1. But more competitive…. that is the reason why so many PhD students are still looking for a postdoc position in those “large labs”. We all need to take sometime to think the consequences of this change, if there is any negative impact for our science, and for those PIs with more than three R01s.

      2. Everyone should look at the actual data that was presented by Mike Lauer in January. As mentioned above, they would not be accepted by most figures journals i.e. the stats used are weak at best. Also, they are graphed on a log scale and not shown on a linear scale. This is simply disingenuous at least, maybe even deceitful. I hope someone with a good informatics background takes a look at them. My guess is that there is a good relationship between funding and publications (only criteria NIH has scored) than presented. Also, this metric only looks ant number of pubs…shouldn’t a Cell, Science, or Nature paper count for a bit more than a JBC paper?

        1. Yes, Cell, Nature, Science papers should count more. For better or worse, it’s a mark of top quality of research. Certainly, there is quality research that doesn’t get published in those journals but it seems that people are taking it too far and are demanding “participation trophies”. Political correctness is running amok at the NIH.

  18. Good idea. However there should be similar rules for NIH intramural PIs as well. Most of the Intramural labs have their per annum funding in millions, yet there productivity is no where close to extramural labs. There should be a ceiling for them, otherwise it will be more like discrimination against extramural labs.

  19. From Nature: “Collins expects the policy to free up 1,600 new awards while affecting only 6% of NIH-funded investigators whose current grants would exceed the new limit.”

    This should be 1,600 new awards in 5 years, or 1600/5=320 new awards in every year, right?

  20. Theoretically, this can be a good thing, but there are a lot of implementational details that need to be addressed.

    First of all, why doesn’t NIH just use “effort” as it was intended? If program thinks a person is spread too thin across grants, then make them put more effort on the grants. That system is already in place. This new system is just a hacked way of approximating effort!

    Second, are there going to be multiple levels for multiple ages? There’s a big difference between a new lab that really can’t handle much more than 1 R01 and a full professor with experience running multiple R01s.

    Third, what really really really needs to be addressed is the issue of continuation, not just providing for more people. What needs to be addressed is a mechanism so that if a lab is running well with 1 R01, then that lab can know, well in advance of the ending of that R01, that the R01 will be renewed. Otherwise, labs cannot run safely on 1 R01. A significant gap in funding can destroy a lab because of the loss of in-house expertise.

    Fourth, how are non-R01 grants going to be dealt with? For example, are T32s going to count against you? Is having your students getting NRSAs going to count for or against you? What about private money? What about collaborations? Is each collaboration going to be negotiated as to the level of involvement or are all collaborations going to be simply divided? Is there a difference between a co-PI (on a multi-PI grant) and co-I (who is just provided “effort”)?
    [Why aren’t we just doing effort, again?]

    1. Your point 3: I agree that should be the top concern. Excellent scientists are not getting tenure, and small labs are closing because they can go from “fine” to “zero” when a grant ends. Of COURSE they can’t compete with big labs in study section. It is not a fair comparison. A single R01 pays half my salary, 1-2 people in the lab, and mice. That’s IT. Two R01s now funds what one R01 used to pay for in the year 2000.

    2. Excellent points….the risk of losing an NIH grant is increasing as the rate of NIH grant renewals is decreasing…

  21. It will be interesting to see if this is a problem that is about to solve itself. NIH has done a terrible job of accounting for the demographics of the baby boom. Many of the concerns about investigator age and lab size are due to the demographic bubble that is only just now reaching retirement age. The number of R01s in a given lab are (of course) highly related to seniority. It is, in fact, highly unlikely that the overbulk of many-R01 labs will continue much longer as the number of senior labs craters. The baby boom scientists are starting to retire and there are very few in the next generation of just-becoming-senior scientists. It will be very interesting to see how the academic system changes when we no longer have that bubble of baby boomers, and instead have a dearth of senior scientists from “Generation X”. I am wondering how much of a real problem this is going to be in another 5 years. This has been a problem for many years, particularly for “Generation X”, but now the problem may well solve itself.

    1. My thoughts exactly; is this just another example of Baby Boomer upper management leaving a departing gift for their Gen-Y children (born 1981-1999) in their first years of assistant professorship? So Gen-X has the rules changed to restrict the benefits of experience or recognition just as they exit from their own assistant professorships?

    2. The Gen-Xers I know are increasingly leaving academic research due to the abject failure of NIH to pay any attention to them. This post is literally the first time I have heard us discussed.

      I am personally giving up my faculty job rather than see my research program implode. Granted I landed something well worth leaving for, but had I not, my remaining academic career would have been dismal.

    3. Most of the baby boomers I know and work with are not retiring. Therein lies the real problem. If you could move their demographic toward retirement, it would help tremendously…but everyone was very opposed to the NIH proposal to coax older investigators to retire by giving them “closure” grants. Still though, that would be the easiest way to free up dollars.

  22. Although I agree with the principle, it is unlikely to produce the desired end result without including the following :
    1) significant cuts to the bloated, comparatively un-competitive intramural program. I have reviewed 4 or so in the last 10’years….There are many excellent, productive scientists in the intramural programs, but there are too many who get 2-3 R01 equivalents with very output that would not survive in the ‘real world’.
    2) increased ratio of total $$ spent on R01s relative to expensive, inefficient consortia
    3) an overhaul of criteria for clinical trials, especially with respect to appropriate budgets
    ***4) a rational assessment of the criteria used to identify researchers that have ‘excessive’ funding. Number of grants, or even total $$$, are flawed metrics unless other mitigating factors are considered. This includes soft vs hard salary support for the PI, and cost of living/salaries, which vary widely across the country, and are currently completely ignored by NIH when deciding on funding levels (e.g. everyone gets a 20% cut on new grants, which has greater impact on those with no choice but to submit higher budgets).

  23. For starters NIH may require salary matching for PIs from universities and private research institutions that have been relying on these army of non-tenured ‘research’ faculty whose job is to crank out one grant proposal after another to keep themselves employed. I would like to see a statistic on how much of the NIH salary support is going to PIs, co-PIs and other senior scientists, as oppose to supporting post-docs, technicians, and research scientists.

    1. I really think the most efficient way at this moment is to reduce “indirects” and ask home institutions to take some financial responsibilities for their faculty members. If 5-10% of indirects reduced for each R01, we can have at least additional 1000 R01s annually. Then we could use $$$ for new investigators…

    2. This is an excellent point! I really really hope that the NIH pays more attention to what is just said in here. I am myself one of those non-tenured research faculty at one of Harvard’s affiliated hospitals and I see many colleagues around me who are 100% on grant money with $0 contribution toward their salaries from the hospital (which by the way, makes an enormous profit). I see first hand how 80% of many PIs effort is dedicated to submit one grant after the other merely to stay employed. Forcing big research institutes to be more responsible to support their faculty is definitely something that needs to be seriously considered by the NIH.

  24. I find it disturbing that no discussion of non-tenure track investigators has occurred. Many fairly competitive biomedical sciences Ph.D.s who might have found ladder track positions in past years now live in the twilight world of soft money, with partial support (if they are lucky) coming from established, older, tenured investigators with whom they collaborate, or for whom they act as a sort of surrogate lab head; taking on many aspects of running a research group, including the writing of grants, the mentoring of graduate students. These individuals in general must self fund much of their salary support, and thus frequently need to get themselves (or more commonly, help the senior PI get) at least one R01 to cover their salary and a portion of their research support. The discussion so far completely ignores where the surplus Ph.Ds have gone (other than leaving academia, which is also common). One of the places they go is into the labs of ladder track PIs. This saves some young investigators, but it can be hard to move from one of these types of hybrid/surrogate PI positions into an independent job. Also, a traditional salary from the institution, even in so-called tenure track jobs, has been waning and is largely gone at many medical schools. Those researchers must fund the entirety of their salary support. Realistically, this means they must get many more grants to survive than investigators with a traditional university salary.
    This proposed redistribution will do nothing for those on soft money, and nothing for the young investigators, with titles like “research professor, adjunct assistant professor, research scientist”, residing inside the labs of more established investigators.

  25. I agree that there is a demographic distortion involved in terms of the baby boom, augmented by the postwar drive to expand US science, so the problem may lessen. However there are some other aspects not all of which are easily controlled but need consideration

    The indirect costs rate: this mainly operates to fund university costs only partly related to the scientific work of the grant. The wide variation in IC needs to be examined. IC encourages the explosion of soft money employees at universities. Both NIH and academia need to reconsider this.

    The special emphasis campaigns — wars on cancer and drug addiction, decades of the brain and other targets, autism etc — what has been the value of these focused programs? Understood that there is political and outside pressure for these activities but they drain funding for the core of discovery science and are too often premature since our knowledge is too limited to take advantage of special funding. The human genome and microbiome projects were overhyped and sooner or later the public and our paymasters will realize this.

    Examine some of the funding strategies used in the EU. Attention is paid to creating cross institutional groups with geographic diversity and young investigators

  26. As a young researcher in the biomedical field, I applaud the NIH for striving to address these issues that clearly exist. However, I think they fail to address one of the worst-kept secrets about the grant review process, which directly contributes to the disparity in funding between senior and early-career researchers.

    While policies may vary by institute (I’m unsure) in the institute to which I submit my work, the review panel is comprised solely of researchers that have successfully received NIH R01 level funding. Consequently, an environment is created where the individuals that are reviewing new applications are the ones that have the highest motivation to ensure that new work does not get funded, because it will take money away from their own, already established research. This further encourages an “old-boys club” mentality that is easily abused through back room politicking and horse-trading among reviewers. This is most problematic and evident in smaller institutes/fields, where the number of senior researchers with existing/prior NIH funding that serve on panels is relatively small. This is why new researchers are consistently counseled to “get out an get to know the experts in the field.” It’s an implicit admission that its often more about who you know, than what you know. This is analogous to the fox guarding the hen house. In line with this (but only my own personal experience), I have had various conversations with senior researchers that have served on these review panels who lamented that “Grants proposed by investigator X are poorly thought out, and justified. However, everyone on the panel knows investigator X, they are well liked, and everyone knows that ultimately the work that will be performed will be satisfactory, so the proposals by investigator X have an extremely high funding rate.”

    Anecdotally, as a your investigator, I can look at my own career and see the effects of this played out. I have been very successful receiving funding from governmental agencies other than the NIH (e.g. NSF). In fact, dating back 10 years to my initial pursuit of funding as a graduate student, every grant proposalI have submitted to the NSF, which utilizes an ad hoc selection of external reviewers for grants and so is less likely to produce a protectionist cluster of senior researchers who vigorously defend their turf, has ultimately been funded. In that same time frame, with a similar number of submissions, I have had none funded through the NIH, and have received comments back on grant submissions such as “Questions arise about the suitability of funding due to the lack of prior experience of the researcher in managing an award from the NIH.” This, despite having a prior history of managing awards from other governmental agencies, and being denoted as having “young-investigator status”; the precise policy that was enacted by the NIH to combat this mentality.

    Clearly, as acknowledged in the article, existing policies to support meritorious research by early-stage investigators are not working. However, until fundamental issues that arise within the selection of the review panel are addressed, I see the proposed policies having little effect on the overall landscape of funding at the NIH.

    1. Ding, ding, ding!

      This is how the enforce their own views about a field, never letting it evolve with the ideas of younger people. Modern approaches are a threat to their established techniques, so they shut innovators out of the funding pipeline. Young people with the correct “pedigree” are favored over upstarts with different ways of approaching science. In my case, this has decimated my career, and I’ve taken a position outside of academia rather than continue to fight the NIH establishment.

  27. So NIH will not be in the business of funding the best science anymore !? How will that keep our country at the forefront of discovery and translation? Very disturbed by this socialistic trend. Your productivity measurements never take into account training – larger labs can train more people and in 20 years those extra trainees may contribute greatly – leaders of tomorrow usually come out of leaders laboratories – now you want everyone to be equal which supports mediocrity ad parity just like in the NFL, which is a horrible product. Please re-think this

    1. You are assuming that the impact scores accurately identify the “best science” and that there is a meaningful difference between percentiles 10 to 20. Often the top scores in a study section reflect seniority/name recognition/number of papers published/bandwagon science rather than the BEST science. Also, the biggest discoveries are often accidental, not what was planned in the sub-aims.

      1. JDyer you are absolutely right. Wallyjoe, the mentoring thing is bogus. I never learned anything from my “mentors”, who were quite well-recognized names in their fields. If you don’t have ideas, the best mentoring in the world is not going to make one a good scientist. Science is not a job or a notion that you learn from someone, because often the mentors contaminate their advice with their own ideas, which may be flawed. Science is something that one learns individually. The lab of a mentor is just a place where one can express their creativity. Creativity is not something that can be taught. What I learned was from other more senior post-docs (techniques) and by my own reading (knowledge and ideas). In fact I found it difficult to convince my own mentors of some good ideas, which eventually proved right. So, “experience” does not count for much in my book, although I can be considered experienced myself. Very often, the more “experienced” people are convinced that they know everything and become a drag against getting out new ideas. The more “established” a scientist is, the more they are afraid of taking risks and possible be proven wrong. But that is not how science should work. Galileo risked being burned for his ideas.

        1. Gabriel: your own experience of “not learning anything” from your mentors is uniquely personal. I do not know your situation: whether the mentors you chose were not willing to mentor or you were resistant to getting mentored since you already were at a level of intellect that did not require mentoring. But your blanket statement that productive and experienced mentors do not contribute to the future leaders and next generation of scientific workforce is not rooted in reality. Where did today’s generation of successful investigators come from (and I am not talking about top 6% rather productive, reasonably funded mid career investigators)? Where the tools to conceive an idea and then reducing that to practice come from? Training and mentoring that occurs in large and productive labs can not simply be ignored. I am very sympathetic to the plight of new/ESI investigators and there are extra-policies geared at ensuring success of these individuals (e.g. a 10 percentile bonus at certain NIH institutes), but “spreading out” the existing monies just to ensure a larger funded pool of scientists may look good at government achievement boasts but never good for science. If NIH is really interested in increasing scientific work force and new discovery, government needs to boost the budget, cut off wasteful and fanciful initiatives (such as RFAs that tend to be non-renewable), make fixed and reasonable F&A costs across the country and set aside a pool of money that is exclusively used for funding new investigators/single RO1 investigators. Socialism does not work in long term, it breeds mediocrity and false sense of achievement since getting first RO1 never guarantees the future productivity of an investigator. Penalizing an investigator who build up programs from that first RO1 they got and continued to run successful programs is no way to increase the output of biomedical research in America. It would rather be detrimental.

          1. @kishraj My experience does not seem to be that “unique”, as the shortage of jobs forced me to go through several post-doc positions, including a couple at Harvard. All the labs were pretty much the same. The “mentor” had a general line of research, asked me to study a general problem (e.g. a disease or a molecular pathway) and the rest was entirely up to me. I learned the techniques on my own by trial and error, or from other post-docs, or from published protocols. The knowledge in the field I learned by myself, by reading lots of papers (and I changed labs and fields 3 times). I tested several ideas, some worked, some didn’t wasting a lot of time on that. None of the mentors ever showed up in the lab. They asked some questions about the results, occasionally suggested more experiments. We talked once a week, sometimes less. The discussions around the results suggested to me that the mentors were less knowledgeable on that particular subject than I was, which was normal, since I was studying that subject full time. At my first post-doc, the mentor’s contribution was to write 50-60% of the first paper, less of the next papers. At next post-docs, I wrote 90%-100% of the paper, the mentor made some corrections and was happy to put his name as corresponding author. That was the entire mentorship. Actually there’s more. One of the mentors took one of my ideas and preliminary results and gave them to someone else in the lab (with no ideas and nothing to do) to finish the project, saying that I have to many projects going on and it is better for the lab to get things done faster. In the end I was second author on a paper that was entirely my idea. No guidance, no discussions on what the state of the field is. That was my post-doc experience. In each case, the mentors obtained grants based on my ideas and findings (in one case the mentor even got promoted to full professorship based on a paper that was my idea, which ended up being published in Neuron, has 300+ citations, but with which he disagreed in the beginning), leaving me on my own to struggle to find a job. This happened in 4 labs, so the experience is not that unique. In fact mentorship is done by peers with whom you work daily side by side. But ideas, which are more important, cannot be taught by a mentor. You either have creativity or you don’t. The bottom line is that the claim by big labs that they do a “mentorship job” and they need a lot of money and people for that is bogus. It is a lie, everyone knows it, but everyone claims that it is the real deal. In fact big labs play a probability game. If you have 1-2 post-docs you put all your eggs in one basket and there is a big chance that you come up empty. If you have 15 post-docs, chances are very good that 2-3 of them will produce something of value, which can bring you grants, Nature papers (sometimes with 15-20 authors, because even the people with negative results in the lab must be somehow compensated), and a tenure. But that does not mean that the PI has more merits, or is more “productive” as you say. It simply means that he has more resources and people. I do not see any merits in that, anyone can go great things with a lot of money. It;s very simple. Tell me I am wrong. The real talent is to get great results with little money, each and every time, like I did. Lots of money also generate lots of waste. I know for a fact that in big labs the PI would ask 2-3 people to go after the same idea, creating a competition within the lab, which is both unethical towards the post-docs and wasteful. In smaller labs (or in my case, me with myself) you have to weigh very carefully every step and avoid waste.

  28. Are we really fixing the right problem? Counting number of grants and dollar values? Some of the route of the problem of lower percentiles is more people trying as was noted. Administrators everywhere pushed expansion of their research programs when NIH was doubling – this makes the denominator bigger and the lament now is a consequence of that expansion as much as the top grant hogs. The mandates and initiatives at mentoring, which is more applicably implemented as advising committees have slowed younger investigators and clearly have not fixed the problem. NIH has many more rules now and maybe more senior investigators have the resources and/or experience to navigate them. Large dollar grants are much more likely to go to individuals who have the administrative experience or teams to navigate their own administrations. Many grants are driven by NIH initiatives rather than new ideas, the proposed solution doesn’t touch these issues. The discussion is a bit vague on contracts and U01s, which often have ancillary studies appended to them. These could be mandated to go to younger investigators, but likely won’t spread the wealth, but will give more grants to younger individuals.

    Spreading the wealth, if that is the real goal would be better served and save much more money by eliminating NIH and make these block grants to the states. While that idea might fly in these times, I am not serious. However, if you want to fund more young investigators calculate how many administrators are eating up the monies because of regulatory needs and “accountability” that might be better spent on younger investigators. The current attempt at reallocation via weighted counting of grants are not looking at the entire pool of dollars and funding sources. They are playing in the pond, when the lake is going unrippled. It may well be that older investigators have a disproportionate share of the funded dollars, but relative to the administrative costs and burdens at or from NIH and local Institutions – it pales in comparison. The solution proposed is not likely to fix the problem en masse and could in fact hurt. If some of the over funded investigators actually use their support for younger investigators, the problem may get worse not better because in many Institutions today young investigators won’t be around to become old investigators if they only garner one RO1 and clearly Institutions have demonstrated that they do not have the capacity or willingness to help.

  29. Its not necessarily the number of grants each researcher gets, its the IDCs that go with them. One route to expanding funding to more researchers is to pro rate IDCs based on the number of grants. Getting IDCs on concurrent grants is double-dipping to some extent. Once the “electric bill” is paid, it does not need to paid again. Why not consider providing the full IDC rate for the first grant, reduce the IDCs to 50% for the second grant, to 25% for the third grant, and none for additional grants. A second route would to end the practice of returning some of the IDCs to PIs as an incentive to write more grants – rolling IDCs back into the research is fine, but IDCs to the PI are not funding research. Go back to using IDCs as they were originally intended – to support the research infrastructure.

  30. Looks like multi-PI R01s are given 6 points while single PI R01s are given 7. Under this point system, multi-PI (i.e., team science) would be disincentivized. I have 2 multi-PI R01s but take in total .8 of one R01 in budget across both, however this would still be worth 12 points. If this system was in place, doing multi-PI R01s would compromise how much funding you could bring into your lab. My other PI is in a completely different field than mine and our work together is far more creative than what we could accomplish alone. The system should not discourage transdisciplinary science.

    1. I agree completely with you. Multi-PI grants or project grants often are an interdisciplinary approach to tackle a problem and should not be encouraged by NIH (which it already is in my field). However, the effort and funding per lab is often lower as funding is spread between 2 or more PIs. If a single PI R01 is 7 points and 2 PIs share an R01, I would think that 3 points would be more appropriate.

    2. Agreed, this scheme only makes sense if the goal is to reduce the number of multi-PI programs (big or small). An investigator with 25% on two U54 grants would have 1 point to spare to raise the other half of their salary (assuming soft money).

    3. I totally agree. The proposed point system of 7 points for single PI grants and 6 points per PI for multi PI grants is very damaging for collaborative science, as well as investigators doing interdisciplinary research (e.g. computational biologists), who often write multi PI grants and need the MPI recognition on such collaborative grants to advance their careers at their home institutions.

    4. Indeed, the current GCI point calculation would strongly disincentivize team science. Hopefully, this is not the intention (and the point calculation will be adjusted accordingly).

  31. One thing that absolutely needs to be considered by the NIH prior to implementation of this new rule is the following. If an investigator obtains a fundable score for a 4th R01, would that PI be able to delay the start of the new R01 until one of his/her other R01s end (if one is ending in 1 year or less)? Alternatively, would that PI be allowed to decide whether he/she wants to end one funded R01 prematurely to accept the new R01?

    1. An investigator should not have 4 R01s. Period. Two R01s is plenty. That is why funding is scarce, because of some hoarders. At 4 R01s, one should start their own company, not expect to be funded from public funds.

      1. Having more than 3 R01s is not hoarding. Those scientists had to write up four different ideas, the ideas had to be judged by their peers (usually more than once), and the ideas had to score below 10% to be funded. This is not easy to do. These scientists should not be penalized for having four great ideas and should be able to pursue NIH to fund them.

        1. @Chris When money is so scarce, it IS hoarding. I know that many “top” scientists are arrogant enough to think that they are the only ones having great ideas. However, FYI a lot of people have great ideas, but cannot get the money to implement them. So, some limits must be drawn. If your ideas were so great, you should have started your own company by now. I did, but because I could not get through the system of connections that governs the NIH reviews, so I started a company to get private money. It is a shame, but I had no choice because of hoarders. If your ideas are so great, you could start collaborating with others interested in them, who have their own funding, if 2 R01s are not enough for you. Why should you get 4 R01s and make others work for you? Why should that money not go directly to the people who do the work? I can bet that some of those 4 ideas are in fact based on the findings of the people in your lab. They should get that money, not you.

          1. You are making a lot of incorrect assumptions that are leading you to the incorrect conclusion that having 4 R01s is hoarding.
            1. Collaborations are of course involved.
            2. The purpose of scientific investigation is not to start companies.
            3. Top scientists do not think they are the only ones who have great ideas.
            4. It is not appropriate to give students or beginning postdocs R01 grants. Science is not an individual pursuit, but instead a team effort that requires funding to do.
            5. A PI cannot make someone work for them. Scientists at all levels choose to work for a PI.

            Great grants vetted by study sections scoring in the fundable range should be funded regardless of the number of grants the PI already has. Stifling the ability of scientists to fund great ideas is never a good thing.

          2. @Steve I am not making any assumptions, I am just describing my own experience, not my imagination. I have been essentially working independently from an early stage of my career, with minimal input from the PI. However, the PI took all the merits (and the money). I am not saying to give R01s to post docs in the first 4-5 years. But after that, they should be able to get one (or a smaller grant, designed for an individual researcher and maybe a technician). But a PI hoarding 4-5 R01s automatically leads to a concentration of funds, pulling the blanket off of everyone else. So people end up having to work for someone, if they can’t get grants on their own. This is what I meant. It is not that they CHOOSE to work for you, it is that they have NO OTHER CHOICE. So they end up working for a PI who gets the merits for the ideas coming from people under him, without having much merit in generating those ideas. And the PI also undeservedly gets the grant money for projects where they had minimal contributions. Smaller grants going to a lot of people would be much more productive than large grants going to a few “pharaohs”. Of course I realize that “big name” professors are up in arms about this because it threatens their status, their monopoly on good publications, and possibly their tenure and income. Very few people willingly give up their advantage. However I hope that NIH officials will consider public interest above the petty interests of “established” researchers.

  32. I suppose if NIH wishes to take on a new experiment to limit investigators funding, then they are entitled to do it (even if they want to call it “Grant Support “index”). However, the statement “we will monitor, on a trans-agency basis, investigators’ Grant Support Index” is chilling. Why should NIH care if an investigator has some funding for completely unrelated, non-overlapping science from a different agency? In my case I have some grants in materials science, completely unrelated to biomedical research. I can produce any numerical measure NIH administrators might want to see that I have been highly productive with their funding (and with all of my funding). The idea that NIH might penalize a highly successful investigator who as some funding from sources other than NIH is frankly alarming. If NIH does this, it will do nothing more than push certain broad and creative investigators away from biomedical research, and/or cause a ripple effect in that those investigators may instead only apply for NIH funding and not participate in other research in the national interest. If NIH wants to ration out their own dollars then so be it, but penalizing researchers for being successful in other funding avenues is entirely unreasonable.

    By the way, I hope NIH might instead have the courage to limit the total number of months of faculty salary charged to grants from all NIH sources to something like four months, rather than the unsustainable level it is now, which can be as high as twelve months. This would be a real change for the better and help to alleviate the need for certain faculty to write so many NIH grants to even pay their own salary.

  33. Thank you. This step is long overdue and will benefit a lot of scientists, especially more junior investigators. This is a big step in the right direction. Kudos to the NIH for realizing this is a problem and for taking aggressive steps to solve it. This will yield big dividends in the long-term.

  34. I’m curious as to why the discussion was about capping everything related to the actual work of research…but nothing about capping overhead. One study not mentioned found that as students population increased 18%, faculty went up 10%. And admin (new deans, chancellors and staff whose sole purpose is often advocating for more funds in this hyper competitive market), increased 68%. DARPA and others have begun to limit the overhead and indirect rates that may be applied in order to ensure the science is the focus. Some have noted that this bloating of bureaucracy was an unintended side effect of the increased research funding in the past, and has contributed to the excessive tuition hikes to ensure it is maintained. Why not add this to the mix so we are funding science, at all levels, not just adding to bloated bureaucracies that serve neither students nor researchers?

  35. My primary and continuing concern about the whole system is that the NIH does not regulate that the institutions have some skin in the game as to faculty employment. The main pressure on faculty comes from the fact that many Universities/Hospitals are happy to take all the indirects and tell faulty get up to 90% of your dollars from grants but make no tangible job commitment to their faculty. The most important element to productivity is stable employment. If the institution wants faculty to have their salary on grants they need to be mandated to make some in kind commitment to the faculty to get those research dollars.

    I like these idea or capping numbers of RO1. I have never had more than 3 at any one time. I am concerned however it is less about age however and more about a few big institutions and supposed leaders in the field getting a loins share of RO1s. My fear as a senior investigator in my early sixties is that institutional and scientific memory is lose if we focus solely on younger and mid level peoples. This is related to maturity of understanding your field and having a broader picture of your field of study and its research problems. I know that my understanding of my field of study is immensely different now than I had ten or particularity 20 years ago. I believe this is reflected also in having senior investigators on review committees at NIH.

    I take a lot of time and effort to mentor my post docs and junior faculty and take pride and feel it is essential to the next generation to do this but their is no value added to this effort. I would suggest maybe having a new grant category that allows junior people to be paired with senior investigators reversing the roles of PI and collaborator. Also as research has become more multidisciplinary there has to be value added for high quality multi -investigator grants and these will need to be assigned a different point value than a single RO1. this may also have value in coupling junior with senior investigators.

    1. I agree. The best way forward with ‘stabilizing’ the biomedical research workforce is to make sure their institutions support and commit to the people, not see them as disposal assets. Capping the indirects, or tiering it down to amount of funding institutions receive. How about removing F&A on salary components all together? Both of these would increase the amount of money directly goes to supporting the faculty, scientist and the research as oppose to supporting the institutions.

  36. I agree with Steve Kron. The simplest way to ensure that the best science is still funded and not unduly handicap a truly exceptional scientist that is somehow able to keep at the forefront of multiple directions, adjusting the payline so that the first 1-2 R01s have an easier line compared to the subsequent ones would seem to be a good solution.

    “I am not sure why a typical handicapping approach cannot be used where getting a third, fourth or fifth R01 is successively harder, like via a sliding scale with respect to payline, so that colleagues can feel the “best science” is still getting funded?

    If the payline for a first/single R01 PI was no worse than 20th percentile but it took a 2nd percentile (likely the best grant reviewed by the panel) to get a PI’s fifth R01, I don’t think there would be as much concern about diminishing returns. In turn, proposals for a fourth or fifth R01 would only be submitted when a PI really had something outstanding to propose.”

  37. How about capping indirect expenditures?. After all, tax rates go up with income–shouldn’t indirect cost rates be scaled back in proportion to the amount of endowment at any given institution?

  38. While I certainly support a system that awards grants according to productivity, I’m against the proposed scoring scheme due to multiple reasons:
    there is a huge (2-3 fold) regional differences in institutional support and salaries in the USA, institutions that require 100% self-funding (including own salary, which at full professor level is $200K an above) and minimum postdoc salary $50,00 and above (~$70,00 with benefits), thus, such policy would disproportionately disadvantage investigators in large metropolitan areas, which are actually the most productive areas and best for training.
    spreading grants across investigators in the name of “fairness” w/o considering scientific output supports mediocracy and not innovative science that’s usually riskier and more expensive. It’s not just the number, but the impact of publications that matter. Bigger labs may publish fewer papers, but they tend to be large higher impact studies.
    Such strict limits will disproportionately discourage women even more to ever enter an academic science career – the more competitive the environment is, the more women will leave and we are already pretty bad with supporting women.
    Instead of such mindless scoring scheme, I think individual evaluation should be used and other ways to decrease the number of “senior” investigators taking away money from younger people: how about proposing a limit to NIH grant money above age 65 or age 70 if the issue really is that large labs run by older PIs get disproportionately more funding?
    Also, how about stricter selection at earlier career stages like graduate school? Currently ~50% of grad students even at top schools do not want to pursue a scientific research career. There is a large drop out at postdocs also, which is fine in some ways. In my view, people who are innovative, committed to success in academic science and have track record to prove this should have no limit on the grant funding they can get, they can set their own limits of what they can and want to achieve. In fact, we want to encourage these investigators to train the next generation of scientist, so that we can maintain high quality innovative science.

    1. I agree with many points you make but I totally disagree with the “unlimited funding”. This is ridiculous. Not even a Nobel laureate should have unlimited money. Regarding educating new generations of scientists, this is also un-substantiated. Firstly, great scientists can be lousy educators, as in my case. I worked for a MacArthur and Howard Hughes laureate and did not learn much, so I ended up switching labs. And vice-versa, mediocre scientists can be excellent educators. Also, you cannot teach having ideas. You either have the motivation and the ideas to back it up, or you don’t. In my case, the PIs I worked for (at several top 25 universities, including Harvard) did not teach me anything. The techniques I learned from other, more senior people in the lab, and the ideas I provided for myself, sometimes against the ideas of the PI I worked for, and it turned out I was right. However, the PIs got most of the merits (and the money) based on my own ideas, while I have a hard time getting a position or a grant. So that is not a good argument. Secondly, most researchers have a peak moment when they find something meaningful, after which they survive on the fame they created for themselves, posing as “experts” but not creating anything else of value. Unlimited funding does not lead to unlimited progress, in fact it is the contrary. Even the most brilliant of minds cannot work on more than 3 projects at the same time and claim that they put in each project the best effort they can. If these “geniuses” have too many ideas, they should collaborate and pool resources with others. This is how it used to be, and I worked out fine.

    2. I agree….this policy does not take into consideration researchers in large metropolitan areas where the costs of living are very high. We now have several special funding mechanisms for researchers in the states in the bottom half in terms of NIH funding. Why not special supplements for researchers in cities with high costs of living?? NIH, why haven’t you addressed this major issue???

  39. Wars for funding among people in different stages of their life do not sound good.The best approach to the scientist and engineer supply pipeline involves increased training for entrepreneurship and increasing opportunities for people trained in science to meet others who can complement their technical backgrounds in order to start new viable ventures.

  40. This initiative has come at an apt time. With more and more young and intelligent PhD’s jumping ship to other industries, I just hope that it is not too late in the game. Couple of other considerations:

    1. Cap indirect costs, similar to Gates foundation
    2. Limit PI salaries to intramural levels. Anything in excess to be borne by the universities
    3. For Multi-PI awards, adjust points system to reflect level of efforts

    1. If one caps the indirect cost rate, how are those costs paid?

      Foundations enjoy the infrastructure and support supplied by NIH indirect costs (F&A) because they are currently riding taxpayer funds; like it or not, Gates would not have a performance site if NIH didn’t largely build and support it. If NIH doesn’t pay, then the idea of supporting research at a university becomes insolvent, making research a low priority, loss leader.

      Bill Gates may be a shrewd businessman, but to suggest we can ALL magically pass the costs somewhere else, is nonsense. There’s a reason for F&A (indirect costs), learn why… find out what rent and electricity cost for your research lab, if your equipment gets serviced, and please ask your grant support administrator if they really need a salary.

  41. The dialogue between scientists and funders of science is missing an important stakeholder: the administrators of educational institutions that host the research centers/investigators. We seem to be heading in a direction in which 100% soft money centers will no longer be viable. Yet, many colleges and universities cling to a longstanding formula of limited to no investment on their part in research centers, beyond what the government provides in indirect costs. The researcher is caught in the middle between a government that understandably doesn’t have enough funds to support all the researchers seeking funding, and their research institutions, whose so-called “investment” in that researcher is all too often limited to what that researcher him/herself has brought to the institution. It will not take many years of partial funding for early career and established “soft money” investigators alike to pursue alternative career options.

    1. I agree that the lifeline of support from NIH and other agencies has created an addiction to soft money operations that is not sustainable and probably not advisable in terms of ensuring quality research. As well, the IC support like heroin has had to be amplified over the years to levels that do not seem to be justified.

    2. I think that has already started. It happened to me. With fewer grants, the indirects are less, and departments have less money. With a poor investments and stock market crashes, fewer tenured scientists are retiring because, (as one PI told me) “my retirement fund took a hit and it’s not as high as I want it to be so I’ll keep working because I’m tenured and they have to pay me- even without a grant.” So departments have less money but they have an increasing financial responsibility to these tenured faculty members who want to retire but don’t feel like they can. They have no responsibilities to invest in the new career scientists (because they are untenured), so…. (as my grandmother would say) “they cut their nose off, despite their face”.

  42. Overall, I think this is overdo, but there is a lot to be done to the proposal.

    1) It’s crazy to count T32s against a PI’s limit. T32s are a thankless time sink, and offer no salary support. No one will agree to direct them if their own science potential is made to suffer.

    2) Counting multi-PI grants (co-PIs) as full charges against both PIs is unfair. The points should be distributed based on % effort.

    3) Soft money positions do exist, and eliminating salary support means a pay cut or job loss. Whether we like it or not, those institutions can’t just pick up the PI’s salary. This is a thorny problem and will require a transition process.

    4) Kron’s suggestion of a rolling payline for additional grants has merit. That would answer some of the concerns about truly exceptional science not being funded, while making sure that the money is spread around. Many labs function happily on a single R01.

    1. other fields — notably the humanities — have had to recognize and adjust to an oversupply of highly trained scholars, by reducing the intake of doctoral students. NIH should consider reducing training grants which are mostly an nconvenience or a holding tank for underpaid research assistants

  43. Given the data, this is an eminently sensible idea. It will promote better science, more productivity, more investigators, and more stable careers.

    The proposed limit of around 3 R01s per PI is reasonable. It still allows excellent investigators to run fairly large labs with several postdocs, technicians, and students, for example. 21 points/3 R01s is a very significant amount of funding, and it’s not like this is going to leave some exceptional PIs high and dry. You can do a ton of outstanding research with 3 R01s.

    As other have said, relatively ‘luxurious’ situation of intramural PIs should also be examined in the same way. A large amount of money diverted to a few intramural researches would also appear to be unjustified. Put the same cap intramurally.

    Overall, very nice initiative!

  44. This solution seems over-engineered. Applicants already propose their percent effort for each year on every application So it is not clear why overcommitment cannot be identified from that information. I appreciate that over-committed investigators simply request permission to reduce effort on existing projects when a new application is positioned for funding, but simply looking at – and perhaps placing more emphasis on – existing effort projections would seem to address this problem. With the demographic transition underway and some of the likely adverse effects identified by other commentators, such as SP, this well-intentioned effort to increase the funding distribution seems like the wrong approach.

  45. 1. NIH needs to get rid of large consortia grants and program projects that are wasteful on administrative costs, inefficient, and low-yield in generating return for the $$. In my experience they have not been particularly effective in promoting fruitful collaboration either. The high level of institutional indirect cost is reaching unreasonable levels and should be capped. Shifting these funds to support more investigator initiated R01’s will maximize innovation.

    2. Age- or seniority-based discrimination in funding is unconstitutional; and it is a slippery slope if NIH continues to deviate from a merit-based paradigm. On the other hand, a tiered approach where 2nd and 3rd R01’s require progressively lower percentile scores for funding, regardless of the PI’s age or seniority, is reasonable.

  46. Although the proposed Grant Support Index would increase funding opportunities for early-career scientists employed as faculty or as researchers at institutions or industries, it is less clear if this benefit will transfer to postdocs. I think NIH should consider expanding their K-awards and adjust the postdoc year limits. Given that postdoc experience is surpassing 5 years, I would hope NIH would increase the maximum years allowed to 7. Also, I think guidelines for reviewers needs to include a statement that more than one postdoc experience should not be scored as a weakness.

  47. Finally, an official has figured this out and has the courage to say that the emperor is naked. Instead of funding a limited number of large labs, forcing post-docs to work forever for a few investigators who hoard all the resources, it is time for a more democratic research, where funding approaches 1/3 of applications, maybe more. This can be done by halving current awards and/or by limiting the number of grants to 2, so that the PI can alternate support from one grant to the other, until the grant that approaches the end is renewed. A post-doc stage of 4-5 years should be enough. 10+ years of being a post-doc due to lack of funding and positions is ridiculous and discourages future scientists. A lot of interested people are choosing instead to go into medicine or business, which offer a more secure income. Maintaining a very limited % of awards is guaranteed to encourage a network of old-boys interests, and of their “graduates”, and keep out anyone who is not part of that network. A post-doc who “graduates” from a lab that is not part of the “network” of a famous mentor has minimal chances of obtaining funding. Chances are even less if the research challenges well-established concepts, as in my case. I have been trying to get a grant for 4-5 years, and have been systematically blocked because of competing interests of reviewers in currently existing models, that proved to have significant limitations. Current system is guaranteed to maintain the status-quo and prevent any ground-breaking discovery from becoming mainstream, unless it comes from a well-heeled lab. To avoid subjectivity in grant awards, half of the score needs to come from the rating of the published papers, and I don’t mean the perceived quality of the journal which is subject to the same network of connections, but the number of citations. A similar system is in place in some European countries and it works out fine.

    The same should be applied for filling academic jobs, instead of current “recommendations” which perpetuate a subjective system of networking and connections. Academia should accommodate more research positions than they do, supported by grant funding instead of teaching, but for that, funding needs to be more democratic. In addition, I find that the NIH grant awards are biased towards researchers with MD degrees, who in fact may be less qualified and have less time to allot to research than people who devote their entire activity and interests to research. I have met many cases where MDs with little expertise in a field hire PhD post-docs to do their work. This is fundamentally flawed and unfair.

    Realistically, a PI with more than 5 people in the lab cannot follow everything that is going on. Sometimes you see lab photos with 30+ people, including several administrative personnel, like a small company, which is aberrant. These PIs need to start their own private companies, instead of using public money to fund these de-facto companies. The PI should be directly involved in the experimental part of research and not delegate this “lowly” part of research to an army of post-docs and graduate students. Excuses with PIs having to write grants, give lectures, etc, are bogus. In my career, I met PIs who did Western blots alongside their students, as well as PIs who, once they got tenure and often before that, were never seen in the lab and became a sort of consultant to their students. I have also seen PIs who kept coming to the lab to do experiments on their own even after retiring. PIs who are not directly involved in research are likely to consider their position a money producing business, not a passion, and should be prevented from getting grants. If PIs had fewer people in the lab, they would have to write fewer grants. There is no reason why most of the grants should go to a limited number of PIs with large labs, who then redistribute the funding to their employees, instead of the funding going directly to those who do the actual research. Papers with 10+ authors, because the PIs must reward everyone in the lab, are ridiculous. If the PI has more ideas than they can work on, they should enter collaborations of equals, with others who have independent funding.

      1. What does it work like, can you please be more specific? I have 20+ experience in the field, enough to form an opinion. Sure, the people at an advantage will say everything is fine. But it is not.

        1. …and the people who are successful are supposed to float those that are not? The other comment or was right…this isn’t how research is supposed to be funded.

          1. I do not see why the best use of the time of an experienced, thoughtful, cutting-edge scientist is to clean beakers in the lab.

  48. This was me- early career PI who had gotten her 1st RO1 funded and trying to transition to mid-career. When it came time to get the second RO1- no longer considered an “early career scientist”, the “Sophomore Slump” hit. Several attempts at a new RO1 were made. My final revised submission was misdirected to the wrong review committee due to an administration error at NIH. That was the final straw and I had to shut my lab. I probably could have kept my lab open through small private grants, but my institution during the time discouraged these applications since they limit the PI salary and the amount of indirects (money going to the university). Now I’m sitting at home, eating gold fish crackers, and trying to find my next direction.

  49. In principal, I agree that funds should be distributed more widely rather than concentrated in the hands of fewer PI’s. But sometimes I find the need to enlist a co-PI who is an expert in a technique that is needed to complete the project but in which I lack facilities or a track record. (I am thinking of RNA-Seq, but the point applies generally). Having established himself as an expert, and built what is essentially a core facility, my co-PI has several collaborators in addition to me. He draws $ as a co-PI from several grants, and is PI without collaborators on others. This is an efficient arrangement, and in the interests of the NIH in that it’s cheaper than funding an F32. But the new policy could make it difficult for me to get a single RO1 because of the expertise and consequent funding of my co-PI. This is even more likely to be the case for new investigators, who may have more need of expert collaborators. If I list my collaborator as co-PI, I will be at a disadvantage due to his other funding. If I don’t list my well funded collaborator as co-PI, my proposal will be criticized because I have not demonstrated expertise in the technique on which I am enlisting his help.
    If this policy is to be instituted, funds received by individuals as co-PI’s lending technical support should be excluded.

  50. I’ve been doing this for almost 20 years. I can tell you that this isn’t going to work and will likely make things worse. Such efforts by NIH are almost always a game of “whack a mole” because while they try to address one problem, they create other problems. I also think that the entire premise is wrong and not backed by robust data i.e. that somehow cutting funding from experienced and well-funded investigators is going to further scientific research. Sounds wrong headed at a basic common sense level, right?!

    Reading all of the comments on this also raised several issues in my mind:

    1) There are already extensive programs to shift resources from more senior investigators to junior investigators at NIH. For example, at NHLBI, junior investigators get a 10% bump in the fundable score they can receive and get funding. With the standard fundable percentile at 14% now, that means junior folks with a score of 24% on their R01s can get funded. That is an incredible advantage.

    2) If anyone looks at the most prestigious journals, many of the papers (maybe most?) come from well funded labs i.e. likely more than 3 R01 equivalents. Do we really want to curtail such potentially ground breaking research?

    3) One question that no one seems to be asking is: how big should the biomedical research workforce really be? Maybe it’s too big and too poorly funded on a per investigator basis. If so, then maybe a bit of downsizing is necessary. This may result in some institutions not receiving much NIH funding.

    4) If research isn’t based on the best science, there is no way around an increase in mediocrity. This will cause damage to the entire scientific community and to our country.

    5) I have served on many study sections, both as a permanent member and ad hoc, and I can say that you DEFINITELY want senior people and not junior folks reviewing your grants. And the senior members go out of their way to try and provide as much deference to young up and coming researchers as possible.

    6) I do agree that indirect cost structures have gotten out of control. This should be addressed by NIH. But it appears that they lack the political will to do so.

    Finally, to those who complain about not getting funding, scientific research is tough. It requires a work ethic that few other endeavors do. It’s not easy-it was never meant to be. This means that some researchers will be funded and others will not. There is no way around that.

    1. I am sorry to say this, but if you have really been “doing this for 20 years”, then you are part of the problem.

        1. I sure want to, but I also want to keep my curiosity, openness, and keep reading and learning new science past 50

      1. ST I totally agree. Some “established” scientists think that hey are the only ones being the bread and butter of science. However, a lot of other scientists could come up with maybe even better research, if they had access to similar funding. When submitting papers or grants, I found often reviewer comments requesting additional experiments that require expensive equipment (e.g confocal microscopy) although that would not bring anything extra in support of the paper/application. They do that just because they may have access to it, but you don’t, so that is a way of shutting out people who do not have access to the same type of fancy equipment. However, people have used normal microscopy for over a century, and most research still remains valid. This is just a small example of what well funded scientists to with their funding. They invest in techniques and then publish a Nature paper based on the fact that they are essentially the only ones with access to that technique. But that is just a nicer wrapping, the content/ideas are often lacking.

        1. My experience in study section service is exactly the same that Productive Scientist describes. The best advocates that junior investigators can get are experienced investigators.
          And the priorities of professional editors at Nature, Science or Cell have little to do with the NIH per-review process. As somebody commented already, that’s not how it works, Gabriel R.

        2. ST and Gabriel: you seem rather bitter, I am sorry that you had it tough and things didn’t work out but please leave this space for productive discussions instead of dismissing someone simply because they have been successful for over 20 years.

          1. Exactly. Two words – Harrison Bergeron. These posters want the NIH to be the Handicapper General. Science will be irreparably harmed by this foolish policy. The quixotic push for diversity at the expense of merit will only lead to a normalization of mediocrity in science.

      2. Fudge. I just realized, I graduated college in 96, and went straight to grad school to work on a project funded by NIDA, from whom i’ve been funded intermittently ever since. My math ain’t what it used to be, but I’m pretty sure I’m part of the problem!

  51. This is a step in the right direction, but permitting one to have 3 RO1’s is still too much.
    I had one NIH grant for almost 50 years and was able to produce a lot on it, including
    starting and leading a new field of cell biology. A university, like my own, with a 25 billion
    dollar endowment should be able to do more for their scientists in the way of bridge funds,
    salary and research support.

    1. Try working with expensive animal models and you will quickly realize that 3 RO1s don’t pay for much. Not everything can be done in vitro or with mathematical models.

  52. This is a step in the right direction, but permitting one to have 3 RO1’s is still too much.I had one NIH grant for almost 50 years and was able to produce a lot on it, including
    starting and leading a new field of cell biology. A university, like my own, with a 25 billion
    dollar endowment should be able to do more for their scientists in the way of bridge funds,
    salary and research support.

          1. Then, if you use them for experiments, you need an entire RO1 simply to maintain your colonies. Before starting paying salaries for your personnel. Before buying any reagents.
            If you have a productive lab based on in vivo experimentation, 2 modular RO1s (typically with administrative cuts) are simply not enough to support a 6-7 people operation, which in my opinion is the critical mass to allow collaborations and enhanced productivity.
            And if you enter into a funding gap, you may need to sacrifice the work of a lifetime.
            The fear of a funding gap is the main reason why productive investigators are constantly applying for funding.
            The rest is bitterness.

          2. @CDO I have no problem with investigators applying for 20 R01s. The problem is that “productive” investigators have created a network of interests which favors their applications over those of other “less productive” investigators. Who’s to say that your application is more valuable than another application coming from a not-so-famous-lab? Who’s to say what the priorities should be in a particular field? These priorities are pretty much set by a small group of investigators who set a particular trend, and if someone else comes up with an idea outside the trend, or God-forbid, an idea that even contradicts the trend, then that outsider will have no chance of funding. And this situation perpetuates itself. Post-docs coming from “brand-name labs” will have better chances of landing a grant than others coming from not-so-famous labs even though their application is not necessarily better. If this group of “trend-setters” and their apprentices hoards a large number of grants, there will be nothing left for everyone else. But I see that this is not one of your concerns. Good for you! Of course, if you have 10 people in your lab you will be more productive than if you have 2 people in your lab. But that does not mean that an investigator with 10 people in the lab is more capable, it simply means that one investigator has more resources than the other, which in itself is not a merit. The real merit is to do more with less. If you have fewer people, you also have fewer expenses with mice, and less waste. Am I right?

          3. Well, everyone knows animals are expensive to keep. However, if you have a problem to support your project by single R01, the way to go is not getting a new R01, but justify yourself stronger to claim more budget than a modular R01(or seek other funding mechanisms), or adjust your research scale fit to it.

  53. Way too many grant applications are funded. The current peer review system often rewards investigators with most papers published which do not translate to public benefit. NIH funded research sould be required and mandated to address real and specific public health problems and clearly describe how the proposed research will help solve the problem (including a timeline or plan for solving the problem).

    FDA uses unbiased internal reviewers for medical product concept evaluation, with safety and usefulness in mind. Implementing similar standards and methods could clearly and very significantly cut the number of funded grant applications, and help NIH direct more funding to projects with high public benefit value, such as pre-clinical and clinical trials that are aimed at solving unmet medical needs. The only and most important criterion should be a clearly identified benefit to the public who funds the research. Fund the science and the feasibility of the proposed solution, not the researcher. I am sympathetic to researchers, I’m one, too, but the taxpayer public does not benefit from increasing the number of researchers. They benefit from research results with clear path to translation towards real life use, including commercial potential or public health actions. Research that goes nowhere except to supporting researcher is just redistribution of wealth from the average folk to a priviledged minority, without any return in sight for the provider of the wealth. NIH (or congress) should significantly increase its support of R41-R44 grants, possibly multiple fold, at the cost of non-translational grants. Productivity can be reasonably measured by outcomes, not published papers, that are in line with the mission of the NIH and the US government.

    1. It’s obvious you don’t know any basic science researchers who do important research but would be irreparably harmed by your requirement.

  54. This will affect young investigators in another way: PIs in large labs with multiple grants will let their non-tenure track researchers be surrogates to apply for grants that would otherwise have been submitted by the PI, with the PI being a co-investigator. It will be interesting to see how study sections respond to this ploy.

    1. This is the first thing the tenured PI in my lab said this morning – I’m going to promote everyone to research faculty and have them be PIs on future grants. Oh and they also get a new investigator bump! Seems like a dangerous and disingenuous way to play the system to me, but there’s nothing stopping him.

    2. This is exactly right. It is already being done to run labs with 30+ scientists, and will be done much much more after this. Administration will promote senior postdocs to adjunct assistant professors as a way of promoting their PIs. There are many ways of gaming this point scoring system, including arrangements where some investigators will be co-investigators with little effort (not on the point system) but still receive large dollar amounts.

  55. The Grant Support Index (GSI) proposed by NIH to limit PI support has major flaws that will deteriorate research in the United States making it less competitive world-wide. In a nut-shell, NIH plans to limit support an individual PI can receive by implementing a GSI cap of 21 points – roughly 3 R01. The fundamental disconnect is that this policy assumes an equal research productivity per funding unit for PIs. In a nutshell – NIH support that a PI receives is counted towards her or his GSI. Therefore the GSI seeks to measure lab support on an individual basis. Research productive is however not assessed on an individual basis but a blanket, average productivity is assumed. Therefore, this approach ignores the high variance in research productivity of PIs. It will have several severe consequences:
    a) It penalize shighly successful, productive laboratories by blanket assessment of research productivity: The amount of funding/grants is assessed on an individual basis (GSI) – research output is not. Highly successful laboratories are curtailed in funding because on average highly supported laboratories have a reduced productivity! To avoid this problem, research output would also need to be assessed on an individual PI basis! Then research output per GSI is used to prioritize funding, ideally as one metric in peer-review – see item (b).
    b) It undermines the peer-review process to prioritize funding: NIH is successful with extramural research as it has a peer-review system in place that seeks to support the best proposals. The GSI will dilute peer-review as ultimately not the best proposals will be funded. As a consequence, overall research productivity might actually suffer. While support is spread wider which, according to NIH statistics would increase productivity, productivity will also shrink as no longer the best research is done. It is difficult to predict how this balances out. In other words, shifting support to more PIs who submitted lower ranking proposals does not guarantee an increase in research productivity overall.
    c) It discourages infrastructure, resource, training grant applications for support: there are several R01 and related mechanisms that support infrastructure, training, and or resources. Grants that provide services to the research community rather than supporting research in a PI’s laboratory will be discouraged if these grants count towards a PI’s GSI. In this case the PI would have to choose between providing a service to the community or supporting her or his own research.
    d) It penalizes large laboratories that focus on biomedical research: Large laboratories that operate primarily from NIH funding versus large laboratories that have a mix of NIH and other sources such as HHMI, DOD, DOE, etc. To avoid this problem, the GSI would have to take all federal and non-federal funding sources into account!
    e) It will hamper collaborative research: PIs will be more hesitant to give subcontracts to collaborators unless their GSI is appropriately reduced. PIs will have to choose between supporting her or his own laboratory members versus another collaborator who is not listed as PI.
    f) The GSI will result in inflated R01 budget requests as PIs will request larger R01s in order to secure the needed support for their research programs for the least GSI points possible.
    g) The GSI will bias support away from laboratories that incur high costs for research: Research that is more expensive in terms of productivity per GSI will get harder as all PIs are measured on the same GSI standard yet incur different costs for technologies and materials needed.
    Overall it appears that the GSI is a flawed approach that is rushed out without considering unintended consequences. Peer-reviewers are in a better position to judge research productive relative to funding support as one metric to rank proposals. The NIH should focus on providing peer-reviewers with individualized metrics for research support (e.g. GSI) and productivity rather than implementing a over-simplified blanket solution on the administrative level.

    1. Amen! these are great arguments – this new metri should be reviewed case-by-case. Many large labs are the thought leaders and highly productive in their publishing, IP and also training (which is ignored by the productivity data).

    2. I agree with you. This is another silly administrative NIH plan that will have consequences as you describe. How about limiting the intramural NIH researchers to the equivalent of one-two RO1s? Most extramural researchers can barely get one RO1 for their labs…..I think that the intramural NIH research is about 13% of the total NIH budget..much too high, in my view.

  56. This is a superb initiative. The biggest reason is that, unlike previous initiatives, it is data driven. The old assumption was that the top 1% of grants and top 1% of investigators was all that matters. There was never evidence for this assumption. We now know that it is better to employ more laboratories with modest funding rather than a few vast lab complexes with lavish funding, because we cannot predict where serendipity will strike.
    Budgets should count in funding decisions. This could come at the institute level. Grants with a great “bang for buck” could be funded preferentially. Study sections with both clinical and basic science applications, for example, will fund a few clinical studies with large budgets and have no funding slots left for bench research. Taking a few lower budget grants “out of turn” rather following a strict percentile ranking could considerably increase the number of funded labs. It would also create a slight incentive to keep budgets low without directly penalizing large expensive projects.

  57. “… and assuring more efficient funding of expensive core facilities.”

    As a Core Director reading this, I am curious about how many of my Core directing brethren there are making such decisions.

    Does the NIH even know what challenges face Core Labs and Core Directors? Has it ever asked in an anonymous way? That is does it realize that the unvarnished truth from the coal-face may differ from the heavily vetted annual report that the receive from PIs and such?

    You speak of retraining graduates. Why can’t the NIH realize that if they insisted on students being trained skills relevant to industry so that grads could transition more easily between academics and industry that would relieve some of the bottleneck.

    Why is it that so few researchers transition back and forth between academics and industry (cross fertilizing as it were). Is it because the coin of the academic realm is your NIH-grant history and H-Index metric? Why should it be that way?

  58. I applaud this initiative, but there are other (and perhaps bigger) problems to NIH funding. To me, the biggest issue is how conservative the NIH study sections have become, dominated by the old guard, with a clear dislike of innovation and risk. THIS BREEDS MEDIOCRACY IN BIOMEDICAL SCIENCE.

    Recently I have been to a study section where 90% of the cohort were M.D.’s who are over 60 years old. I am mid-career, and I was the youngest person in the room. What they considered as innovative to me was archaic work, and what they considered as risky was barely new and innovative.

    1. Amen! The NIH is shooting science in the foot by allowing the old guard to protect their territory at the expense of new ideas.

  59. Preventing great ideas from being funded because the scientist already has 3 R01s, will do just that–prevent great idea from being pursued. Some people are more creative or have more diverse interests than others and should not be penalized for it. Moreover, it is difficult to measure what is gained from a large lab versus a small lab since the returns are not always obvious or immediate. Large labs with more money can do bigger things that usually lead to multiple other projects picked up by smaller labs, but this cannot be measured. Both sized labs can make big discoveries and having diversity in lab size and funding contributes to this.

  60. I would consider another approach too. Very close examination of indirect costs. I would guess that the actual costs do not increase with each additional grant received by a PI due to overlap of resources (e.g., space, utilities, etc.). Yet, each grant awarded gets the full IDC. So for a PI with mulitple grants the institute is getting IDC above the actual increase in costs for additional grants. I think each additional grant should get reduced IDC. This would free up a lot of $.

    1. This is a great idea because it directly addresses a core challenge. A lot of the perverse incentives come not from the researchers in the trenches, but from institutions that directly benefit from indirect costs. After the NIH doubling, too many major academic centers changed their business models away from education and training and instead rely primarily on grant funding for revenue. Institutions set tenure and promotion policies to encourage investigators to bring in as much indirect funding as possible. This is a major problem that needs to be addressed. One option is to limit indirect rates to two concurrent R-01s. Investigators could still receive funding for their next great idea, but the overhead wouldn’t come from the NIH. Another option is to cap the total costs, rather than direct costs. NIH already does this for SBIR/STTR grants. Addressing the indirect rates, and the incentives behind them, would leave more room for the science.

  61. The NIH could easily free up more money for investigator driven R01s by limiting indirect costs to a standard 25% for all. My institution receives about 54% many get more, even up to 100%. Indirect costs are an incentive for institutions to hire people in the hopes that they will not only fund their salaries but generate more cash for the institution.

  62. It is quite likely that the proposed NIH plan was discussed over Starbucks, researched with Google, written with Microsoft, formatted with Adobe, on an Apple/Dell/HP computer. The metrics of the proposed plan would certainly conclude that these companies are excessively large, wasteful and unproductive; it would cap their funding and metaphorically decimate their infrastructure, innovation and visionary leadership. The intangibles of a large company, or lab, ranging from the fertile training environment, to the scale and innovation of research and development, to the visionary thought leadership that can boldly advance a particular field, would be lost with the proposed approach.

    1. Perhaps, in order to fund more scientists across the board, we could allocate fewer NIH resources to planning and disseminating annual overhauls to application forms and instructions, biosketch formats (forcing us to create a new version for every application), and burdensome reporting requirements. And maybe funding decisions could be made and communicated over a span of time shorter than, say, 18 months. That would really facilitate our planning and reduce the number of applications submitted as back-ups.

    2. Very well stated! It would be interesting to hear of examples where redistribution of resources on this scale has led to increased scientific productivity…anyone?!

  63. The problem is that the suggested scoring is catastrophic because it discourages collaborative science and even penalizes training grant directors who do a fantastic job to help their institutions with no benefit at all to their own research program.

    Here is the site with the scoring table: https://nexus.od.nih.gov/all/2017/01/26/research-commitment-index-a-new-tool-for-describing-grant-support/

    The suggested scoring is irresponsible and indicates a lack of understanding how science works. To explain, P50 structural Centers for AIDS research allocate less money to individual investigators than P0-1s yet participants are hit with 11 or 10 points. Both P0-1 and P50 often only bring in ~$150,000-180,000 annually while R01 bring in ~$250,000. Given P50 centers 10/11 points and equalizing P0-1 with R0-1 (7 points each) will discourage these interdisciplinary, innovative and collaborative grants. The analysis of the AIDS structural centers has revealed that they present some of the most productive, most innovative engines of science. Discouraging participation in collaborative efforts will make science less effective, not more effective.

    A limit on dollars per investigators is much fairer. Some scientists are more productive in collaborative settings than others and the NIH should not be allowed to rule how each scientist creates structures that make his/her work most productive. Such limitations will limit the effectiveness of our research.

    One more point, the word “socialism” is already seen on this blog, rightly so. I also came from an East European communist country to the US over 22 years ago. America stands for the freedom and strength of the individual as an engine of productivity that benefits the country as a whole. One rule doesn’t fit all and implementing these rules will drive some of our very best scientists out of this country. It would be much fairer if each individual is assessed and evaluated to see if he/her’s productivity maxes out at ~3 grants or not rather than implementing one rule for all.

  64. What I Don’t like are:
    Caps (3 good ideas Herr Einstein, now you can turn your brain off and stop writing ideas down – they will not be considered)
    Bans (an FOA that bans anyone who has current funding from us)
    Slicing the sausage thinner (yay, everybody get 50K/y to do ground breaking research)
    Complex and inaccurate ‘derivatives’ (like the talked about GSI – in any incarnation)

    What I Do like are:
    MERIT – and allowing everyone to compete and having the best ideas from review have a chance to get funded – so here is a very simple idea: If you have a certain number of active awards and you are being reviewed on another (let’s say your 3rd R01 for example), you have to do a certain % better than the payline to get award (ex. if the payline is 10% then you have to get 8 or 9%) – so simple and makes for a world where someone extraordinary can have 4 or 5 good ideas at the same time and still get funded while addressing the “wealth begats wealth” concern. It mirrors economic 1%-er issues – we love capitalism & hate communism but too much wealth in the hands of too few is a problem and can be addressed through progressive taxes and closing loopholes but we don’t ban people from working.

  65. I would suggest that there be a more fine-grained assignment of GSI points to U54 grants. They are not all equivalent and not all aimed at supporting the PI’s research. For example, NICHD funds the Intellectual and Developmental Disabilities Research Centers. These are almost entirely core based centers where the PI serves as a manager and not a researcher. If the goal of this new plan is to spread the wealth of funding across researchers, then PI’s should only be considered with respect to their research grants and not those grants that provide infrastructure support for the larger community of other investigators. This would be the same for a T32 training grant that does not benefit the PI’s research lab in any way.

    1. Since T32 grants are linked to a PI, it adds free personnel and should therefore be considered within the PI’s cap unless the mechanism is changed completely to provide complete authority to the trainee, including the freedom to move labs and institutions or select multiple mentors.

      1. Just to clarify for KS: PIs on T32 grants are not getting personnel to do their own research but rather organizing a group of trainees who are in the labs of others. Fellowships (F32, F31, F30), on the other hand, do provide a benefit to the lab of the sponsor but the PI is the student or postdoc. T32 PIs should certainly not be penalized for the work they do for the entire community since they get (usually) no salary or research support from those grants. It’s an alphabet soup for sure.

        1. I stand corrected on T32. Thank you. I meant F32, which should be replaced by K awards. I agree that the supported parties should share the responsibility. I agree that a lot of detail has been overlooked and feel that implementation of this process will face opposition from senior investigators in the advisory boards that will be nearly impossible to overcome. A fundamental problem remains that the number of grants rather than dollars involved are being targeted, although I can understand the need to reduce the burden of reviewing a large number of grant applications.

  66. I am generally supportive of the effort to impose limits on NIH grant support but the proposed policy will damage multiuser research centers and core facilities at a time when they are becoming increasingly essential to strong biomedical research. Specifically, the proposed Grant Support Index will make it impossible for a PI to lead their own strong research program and to also lead a multiuser national center.
     
    My particular concerns are the weight placed on center grants relative to a standard R01 or P01 grants, and the disparity in weight placed on different center grants, such as P30 (4pts) and P41 (11pts). These center grants do not support a PIs own research program, but rather support technology development and access to resources, expertise and technology to the broader community. The penalty for being the PI on a grant funding a multiuser national research center will be loss of research grants for that PIs own research program. I strongly believe that NIH should reinforce the synergy between strong research programs and research centers, which are of huge benefit to NIH-funded researchers nationwide. Some adjustment must be made to the Grant Support Index to correct for this damaging consequence of the proposed policy.

  67. I applaud the NIH for acknowledging these long-standing, problematic funding issues and for committing itself to correcting the problem. Great ideas and impactful research can come from anyone, not just certain individuals or institutions. So often, promising scientists don’t get funded if they are not affiliated with more established/prestigious PI’s or labs. It’s a typical case of the “haves” and “have-nots”. The “haves” get more, the “have-nots” get little to nothing. This should not be the case in advancing science and improving the quality of health and life for everyone. The scientific careers that get stunted due to lack of funding could be the very ones that would have found an effective treatment for a disease.

    I think funding caps are a great start and will be instrumental in correcting the problem. After 3 rounds of major funding, well-established labs should have the equipment and resources they need to continue their research. Their institutions, state funding sources, or private benefactors can provide support for student and post doc positions.

  68. Dear Mike,

    I work in liver research field. Many PIs voiced their thoughts on the ‘point system’, and I was asked to provide some feedback. A brief summary is listed below:
    1. Due to the relative high points for R21, less applications will go for R21, more crowded on R01. Not sure it is good for exploratory studies without solid preliminary data for juniors.
    2. P01 and R35 have money about 2-3 R01, but have the same point as R01. The point system won’t do the work. More PPG will go to NIH.
    3. To promote collaboration, MPI should not double count. NIH should limit total $ amount per lab, not # grants, and reduce indirect costs. The problem is every institute reduces PI but increases Deans and put pressure on PI to get more grants and results in current situations.
    4. There will probably be less MPI proposals, which diminishes collaboration.
    5. A PI of P50 could get 11+one sub project 6=17, and may be more, if administration count. If this is correct, then he or she cannot have a R01.
    6. In summary: R01 can be funded by 3, 4, or 5y. R21 is roughly one modular of an R01, is only 1/4 or 1/5 of a regular R01. It is off proportion to assign 5 points for R21 vs 7 points for R01. Same concerns for MPI (depends on how many PIs).

    Overall, the current point system will 1) discourage collaboration MPI grants; 2)discourage innovative/exploratory R21 grants; 3) promote PPG (which in fact concentrate funding within several groups), which does not benefit most PIs.
    Thus, total $ amount per lab, not # grants, is the better way to go.
    Many PIs feel indirect costs should be cut.

    Li

  69. This policy will shine a light on the dark recesses of grant support inside the intramural NIH program. There are examples of PIs, including single NIH IC directors and chiefs, etc getting in the range of 4 to 36 R01 equivalents PER YEAR (i.e. $2M to 9M per year for a single NIH PI), and have zero senior author original research papers (not counting reviews, editorials, perspectives, letters to editor, and editorials) for the years they received NIH support. In answer to the point by Dr. Lauer “analogous steps should be taken with NIH’s Intramural Research Program?” – is now an existential one for the entire Intramural Program. That’s because once it is known, by a published quantitative analysis, how little productivity each NIH intramural investigator has on average per taxpayer dollar, the community and taxpayers will not tolerate this. If NIH does not apply this policy to the intramural program, there will be calls for conflict of interest (e.g. 17 R01 equivalents for NIH Director’s lab in 2016, with zero (0) senior author original research papers in 2016). For this reason, there will be many consequences to this policy, including a shakeout of the Intramural Program.

  70. Whether one agrees with the premises supporting introduction of this policy, one can ask whether it will really result in more awards available to be distributed among the less fortunate. Two strategies that will likely be employed immediately among the 21+ crowd are likely to minimize the actual number of new awards available. First, a likely immediate strategy amongst those with 21+ GSI is to removed themselves as mPIs from all MPI awards. this alone is likely to substantially diminish the number of investigators who are 21+. Second, one can expect that any new (or renewal) competing applications from the 21+ crowd are likely to request 33-50% more funds. If necessary, grants that would be targeted for termination as part of the required plan at each competing submission, would be the smallest R01 that investigator held. One can anticipate a significant increase in the size of grants among the most successful applications as a way to maintain the cashflow while complying with the new policy. As a result of these two strategies alone, it is unlikely, in my opinion, that there will be many slots freed up for additional investigators. There are many other strategies that could be invoked, but even these 2 transparent approaches could doom the new policy’s effectiveness.

  71. The assignment of points to different grant and cooperative agreement mechanisms has some puzzling quirks. For example, R03s, a very small research award, are assigned 4 points. R24s (and I assume U24s which are missing from this list) are also assigned 4 points. yet if one looks at the kind of projects supported by R24/U24s, they include very complex projects with multimillion dollar per year budgets. Even P30 centers are assigned 4 points, even though many of these center grants run P&F programs that award multiple R03 sized projects each year in addition to supporting a number of core activities. On the complex mechanism side, U54s are assigned 11 points. subprojects on multicomponent awards are assigned 6 points. So the PI of a U54 that had multiple components could not be a PI on a subproject and still hold a single R01. Is this really what is desired? I could continue as there are many more examples, but I would suggest that someone with broader experience with the multiple uses of the multitude of NIH award mechanisms take a look at this list.

  72. “No problem can be solved from the same level of consciousness that created it.” —Albert Einstein.

    Solution step one: Establish a NIH review branch with full-time NIH examiners who never apply extramural grant themselves. (Look for Einstein in biology)

    Solution step two: Self achieve and free update the PI’s research results in an open access or NIH access journal database created by NIH. No impact factor, number of paper, nor journal publication is needed for funding decisions. PIs can keep revising their idea and data in the database. The system will record who had first idea, who followed others and who did plagiarism. (Everyone will be interested in posting more ideas rather than more papers)

    Solution step three: The NIH review branch first defines qualified PIs using a relaxed standard (20+ can publish Nature paper in the old time), score PIs based on innovation, importance, quality of the work in the database, and fund multiple small budget modules according to the score. (Innovation can happen in limited budget)

    Solution step four: Start with a pilot number of participating PIs on a voluntary basis until replacing the old system a decade later. (Young scientists join first and status quo joins later)

    Solution step five: Continue to refine steps one to four.

  73. The intentions of the proposed changes are good, but everyone knows that the road to hell is paved with good intentions… There are always unintended consequences when large structural changes are introduced to complex systems… As Peyman Golshani and others pointed out above, Inhibition of collaborative grants is one of the many possible serious negative consequences of the proposed changes. So, how do we move forward? We need to carefully introduce smaller changes that continue to tilt the balance towards smaller labs and more money for young scientists… History has taught again and again that jolting a complex system is almost never the best way to bing about smooth change. We have too much to loose by risking a rough rocky change that ends up by doing more damage than good. What we all want is better use of NIH money for more innovative science. If we are not careful and measured, what we could get instead is a twisted set of incentives where nothing really changes and big names continue to get the big money with surrogates and other equally manipulative mechanisms… Summary: great idea, but it would be foolish not to introduce these changes in a gradual, “let’s see manner” that is designed to avoid perverse unintended changes. Gradual changes would allow NIH to adapt and correct the course if things do not work as planned…

  74. NIH should reduce 10% of the “indirects” for each of their grants immediately, which won’t have any significant negative impact to any small institutes since they have less grant $$$. For Harvard and other larger centers, they also have enough $$$ to cover for the reduction anyway. In this way, you also downsize the “administration”….

  75. This system is going to totally destroy collaborative research. For example, I am a PI on an multi-PI R01 and only receive $80K in direct costs, yet my effort is important for the work. If this is going to be scored as a 6, then what motivation would I have to participate in such a project? This system is going to force people to write very large budget single PI R01s, and justify the budgets with equipment, supplies and personnel. This will incentivize people to not include collaborators that are increasingly playing a very big role. It also doesn’t take into account large grants that will expire a year down the line. If a grant is disallowed because the points are too high, a lab may be left without money a year into the future for example.

  76. two points 1) cutting NIH intramural fundings as NIH did for extramural grants in these years, and 2) reducing 10 to 15% of the “indirect costs” for extramural fundings. These changes will make an immediate and positive impact for science in the nation.

  77. Remember Goodhart’s law: “When a measure becomes a target, it ceases to be a good measure.”

    We already have plenty of measures that limit the number of grants a PI can receive. In the new system the PIs who will succeed are the ones who can play by the new rules and promote their postdocs to research faculty so they can get grants.

  78. NIH is trying to correct a perceived problem by penalizing great scientists. There are huge flaws in how productivity is being measured here–papers are not the only thing big labs generate. The initial impact of implementing such a ridiculous plan will be a reduction in collaborations and training, which has taking NIH a generation to improve. Moreover, scientists are very smart and will figure out away around these limits. As others have already pointed out, there are ways to get around this cap that will take time and energy away from science and just make more work for everyone. Study sections work hard at identifying the best proposals and yes, some excellent proposals do not get funded initially due to the low funding levels. However, persistence usually results in a funded grant. Therefore, please NIH throw this stupid, ill thought out idea away. Do not put restrictions on the number of ideas any one scientist can submit–this will only hurt scientific progress in the US.

  79. Several ideas in the comments on this blog are well worth noting.

    In particular, the idea that there are hard caps is a real problem, but making it harder to get second, third, fourth grants makes more sense. (Cutting the paylines tighter for later grants, loosening them for labs in crisis.)

    However, the link to the RCI activity codes (which I guess are the GSI codes) suggests major penalties for all the things that make the system work – T32 training grants, R25 undergraduate and diversity grants, multi-PI grants, etc. Running a T32 or an R25 is a thankless job of a tremendous amount of work that only gets done because scientists believe in the next generation. If anything, it hurts the PI and should count as negative GSI. (Actually, the idea of increasing GSI could be used for rewarding things like serving on study section, etc.)

    Earlier studies seemed to suggest that $800k/PI was close to optimal on average, but that included money coming into your lab from student grants [NRSAs], T32 fellowships [to the lab the student was in, not the PI on the T32!], and private moneys. If the proposal is to include T32s and R25s as being in the GSI of the training grant PI, then of course there would be a reduction in productivity as a function of dollars, because the papers coming from that T32 are not in the PI’s lab! Similar arguments need to be made about P50, P30, and core grants. We want to encourage those grants, not discourage them.

    There are also major problems with the way the proposal handles multi-PI grants. We have been encouraged for years to start running multi-PI grants, but the numbers seem to suggest GSI won’t track money here. If I’m providing less than one aim’s worth of help to another’s grant, it shouldn’t count almost as much as a full R01!

    As noted by several comments, any hard-coded system is game-able. What we need is a flexible system that expects a distribution of PIs from small “single-R01” labs to large empires. And moreover provides the opportunity for PIs to grow. The GSI system does not appear to be flexible. Instead, it seems designed to force all labs into a single format. There needs to be variable structure based on age and track record. I was not ready to run a large center or a T32 grant 15 years ago, but I am now. 15 years ago, it would have interfered with my R01 productivity. Now, my lab has processes within it that I can run my R01s and a T32 without sacrificing productivity. When I was young, the senior faculty had T32 grants that helped me. Now it is my turn to try to get T32 grants to help my junior faculty. But doing that shouldn’t count against my GSI!

    Mike – Are you listening?

  80. How poorly thought through is this policy? Example: One U54 or P50 PI pulls together many investigators to run cores and/or projects, but his/her only role is to organize and manage, while other investigators run those cores and projects- 11 points. Another U54/P50 investigator runs the center, leads a core, and is PI of a project- 11 points. The former is bringing in other scientists including junior ones, the latter is getting a great deal of funding.
    How about a simpler, less intrusive alternative to the points system? Work within the maximum percent effort allotments already in place, but suggest a percent effort minimum. This is already in place for P50’s and U54s (20%); additional projects could require 10% effort on a P50 or U; R01s could require 15% effort; in the end this would put a de facto cap on the number of grants while allowing those of us on soft money to make their salary.
    But really, this is a policy that will kill multidisciplinary team research; will greatly harm truly exceptional and groundbreaking labs (do we really want someone like Karl Diesserroth to downscale by 75%? No!).
    How about eliminating top-down, huge-scale, hypothesis-free set-asides, and just fund the best science? There is huge waste in many areas. NDAR, a foolish attempt to combine data across sites retrospectively with no scientific controls in place, no hypotheses, and minimal interest in the community, is costing millions each year. I am sure there are many other examples.

  81. I think there are several good ideas here, with which no one would argue, such as the need to grow mid-career scientists and ESI. The inherent conflict here is one where redistribution of grants is determined by an index that will diminish the incentive of established well-funded researchers to go with their best ideas to the NIH. The suggested policy will enforce that after three R01s, no more great ideas can be funded from that PI over a given time period. The key to the NIH grant funding system is a peer review system, which despite its manifold flaws, seems to work reasonably well. Are there options such as escalating the bar for funding incrementally on a sliding scale (such as a score below the 5th percentile) once investigators cross the 3 R01 threshold? No one would argue that a grant with a first percentile score as assessed by an independent peer-review system should be considered seriously for funding as the science proposed is likely to be impactful. The proposed system could have other unwanted side effects such as shadow writing of grants by senior PIs to get great science funded, and this will not necessarily help younger PIs in the long run and will limit the aspirations of talented mid-career scientists as they view the road ahead.
    Whereas principles of distributive justice seem reasonable to implement, it is equally important to nurture a meritocracy that respects peer review scores and does not systematically discriminate on the basis of performance. Clearly a middle path is required that brings together different core principles in play here, rather than a policing system that is driven administratively and over-rides the peer review system.
    It is also worth noting that limiting and controlling the time allocated by senior investigators to science funded by the taxpayers is disturbing. The junior scientists of today are the senior scientists of tomorrow. Why build a system that violates some of the key attractions that draws scientists to this country?
    A fundamental disconnect is a relative lack of discussion of why we are here in the first place! We need to evaluate (in parallel) more effective options for lobbying Congress for more funding in order to nurture the current and next generation of scientists better. Implementing a system that caves in to limited NIH funds and regulates more productive scientists based on a GSI threshold will not address the ‘root cause’ at play here. NIH leadership and the scientific community should partner with each other to educate the public and congress more effectively on the need for additional science funding that could then preferentially be directed towards early career scientists without regulating senior ones; in short, ‘ Expand rather than ration the pie!’

  82. In general I agree with the motivation and goals of this initiative. However, its success depend on the details of the GSI formula. The only available information on the GSI seem to be that it is likely to resemble the “Research Commitment Index”, which is outlined at https://nexus.od.nih.gov/all/2017/01/26/research-commitment-index-a-new-tool-for-describing-grant-support/

    This is a little worrisome. Just to consider RO1 grants alone, here are some obvious problems with the RCI:

    (1) All single-PI RO1 grants are assessed the same in terms of points. This is even though there is tremendous variations among RO1 grants in terms of total direct $, % effort of the PI, number of other faculty-level personnel receiving funds from the grant, and balance between personnel costs and other direct costs.

    (2) As others have pointed out, in the RCI multi-PI RO1 grants are assessed nearly as many points for each PI as a single-PI RO1 grant. This is true even though there are plenty of MPI grants in which 2 or more PIs are splitting a sub-$200K budget.

    Based on the above considerations alone, I would predict that investigators will respond to the GSI by ditching MPI grants altogether in favor of single-PI grants in which collaborators still receive funding, but without being afforded PI status. It will be hard to block this loophole without discouraging collaborative research. Moreover, it can be expected that PIs will shift towards larger budget requests for each grant, since there is no penalty for doing so in terms of GSI points assessment.

    I respectfully hope that this initiative will not be rushed into practice without a thorough vetting (beta-testing?) of the GSI formula.

  83. This is an ill-timed and and ill-conceived idea, more like a knee-jerk response to an anticipated funding cut (which was looming early in the year when this was first introduced). But now, as NIH received a $2B boost, it seems out of place.
    First, the situation of junior, and not so junior faculty is extremely difficult. For many, getting their 1st R01 seems an impossible task. But there are many more fair and realistic ways to address that. Just a few ideas: NIH could set aside some of this new money for new grants submitted by 1st time investigators. The biggest problem for junior investigators is that they lack mentorship and guidance about what is expected from them and what a successful application looks like, and they are expected to compete with established labs. So NIH could instate mentorhsip programs, and make the review system more transparent and fair. Another approach could be the creation of smaller grants, specifically designed for junior applicants, that are maybe 100K/year for 5 years, and one could get several of them, thus minimizing and distributing risk for NIH. Those could then create the basis for R01s in the future. The current ultraconservative grant review system could also be revised, to fund ideas and promise, not overwhelming preliminary data and grantsmanship.
    The idea is also ill-conceived, because it plans to use the same simple idea of wealth redistribution as used by the Bolsheviks. This seems fair at first, when seen from the perspective on an applicant struggling for a first R01, and is a simple populistic measure from the part of NIH. However, dis-incentivizing high performers has not worked before, and is not likely to work. First, there are several loopholes already alluded to by Francis Collins’ letter that well-positioned PIs could use to bypass this rule. So it will primarily affect upwardly mobile, highly successful mid-career investigators. Then, many highly funded labs do actually produce a lot, have large projects, that contribute to science immensely. Some of these projects do require lots of money, like GWAS, large mutagenesis screens, clinical trials. Connecting real productivity in recent years with future funding is a fair way of addressing merit based funding for established investigators. On the other hand, it is true that some highly funded PIs use primarily influence, seniority, past accomplishments, networks of friendly reviewers and program officers, to get this money. But punishing everybody for this is not fair, and will definitely be counterproductive. Instead, it is well within NIH’s power to curb favoritism, make the process of funding transparent and fair, distribute NIH’s “dark money” (that is given to large labs without a real review, based on program announcements posted 2 weeks before the deadline and specifically tailored for 1 lab). And lastly, many of today’s junior investigators will one day be established and applying for their 4th R01, and many of today’s recipients of their 4th R01 were not so long ago struggling to get their first. From their perspective, this measure is not fair.

  84. I generally agree with this idea and feel that its time has come and gone several times. I somehow feel that the idea will be challenged by the best funded and most connected scientists and therefore never get implemented. In addition, there are issues associated with this type of social engineering by the NIH. for example, the push for early stage and young investigators by Dr. Zehruni was part of the cause of this impasse. Early stage investigators have always been pushed in large numbers and generally came from dynamic eminent labs, mentorship support, support from numerous agencies, tons of data from their previous labs and a lot of institutional startup support. However, investigators (even tenured professors) who were once successful, but took risks and lost funding, get no support, are still pushed to cover a huge percentage of salary, and are otherwise punished severely, leaving them in a no win situation. This is a recipe for handing success to the mundane paper pushers who are also glib and connected, while risk takers who quietly pursue important scientific hypotheses are often destroyed. It is important to remain balanced and consider all faculty with $0-250,000 direct support in the last 3 years at a base level and progressively evaluate grants to make them compete with others at the same funding level. By creating a tiered playing field should make competition fair and give the best science the best shot at success. Young investigators will populate the lowest funding tier along with some mid level and senior investigators. Even without a cap, competition will become more severe at the top and the best scientist/managers will still get funded to push the best science without caps.

    It is also important to limit the number and size of grants such as PPGs, which are individually reviewed and therefore not subject to the same competition. Clinical trials and other huge grants could be told to not just compete for NIH dollars, but asked to find matching support from the private sector, which will also carry out its own due diligence, ensure value and find ways to terminate failed trials. Other clinical trials supported by government should be limited to products that the patent driven industry will not support, such as diets and natural products or comparing existing drugs with new ones.

    A third issue relates to the mechanisms by which state institutions obtain funds from the federal government. In particular, several state universities successfully palm off their responsibility to cover faculty salaries to the federal government by capturing bulk of the PI salary on federal grants while still using them for teaching, administration, training and grant writing time. States have conveniently cut back support for their higher education institutions. To avoid this problem, NIH should clearly limit percent salary coverage to 50%.

    NIH should also try and recover all funds from the institutional share of intellectual property-related income based on facilities and administration cost and percent income of tenure/tenure track faculty covered on NIH grants. These funds will allow NIH to expand beyond federal government support.

  85. Question: for the calculation of RCI, does each subproject contribute 6 points? Or, are there specific activity codes where subprojects would be counted? Further, as an example, if a scientist has a U54 where she is the single PI on the award and three of its subprojects. Would this investigator have 11 or 29 points in this scenario?

  86. There are deep issues with the proposed grant strategy. Perhaps the most problematic is that research decisions with this policy will be made by bureaucrats and not by scientists. Imagine that a scientist applies for their fourth grant and receives a better score than someone who has just submitted their second grant. Under the new policy, the less successful grant application is going to be funded. This is absurd. It doesn’t help science, it hinders progress. The mission of NIH is to drive progress, and that means putting all grants on equal footing, and allowing the best to rise to the top. Tying into this issue, is that the entire idea being proposed relies on average behavior of grants. But in quantifying progress, the average is meaningless. What is important are individual contributions. Sometimes a single contribution is worth 10,000 grants. Only a merit-based approach makes sense.

  87. Did the assessment take into account whether the PI is at a college or university vs a research hospital/institute? Professors have to commit serious amounts of time to teaching classes and mentoring undergraduates. PIs at research institutes do not, and thus have significantly more time to commit to their research. They may be able to effectively manage more than 3 R01s because they have much more time to focus fully on their projects rather than their teaching load.

    1. Commit time to teaching vs. commit time to another R01 project vs. commit time to clinical duties, what’s the difference? I’m sure you’re not suggesting that time committed to one R01 project, helps towards another, because we all know we can’t ask the government to pay twice for the same work. Time committed elsewhere (wherever that is), is that much less time available to commit to a particular grant. So in terms of time available, it is immaterial where the time is being committed.

      1. I don’t teach, so I have many more hours per week available to do science. Maybe people who teach can only handle 3 projects because they spend half their time teaching. Since I am not teaching, maybe I have the time to effectively manage 4 or 5 projects.

  88. This is another attempt for NIH to cover its inefficiency at managing funds and maybe even corruption. Check for grants larger than $15,000,000.00 in https://projectreporter.nih.gov/ and you will find ~100 listed. That is about $1.5B , a significant chunk of the $30B NIH budget. This is also the equivalent of ~700 R01s (average R01 is $2M). My opinion: no need to add another layer of bureaucracy at NIH to just count more beans (or points assigned to researchers). Instead, put the largest grants given at once under more scrutiny! Care more about the health of the grant-awarding mechanisms! And stop wasting time on arbitrary rules that only generate more rules and exceptions and more bureaucracy.

    1. Totally agree that NIH should look into large multi-million dollar grants along with limiting R01 awards. I have seen FOA from institutes like NIDDK for single 50 and 30 million dollar awards for establishing clinical trial centers when there is no effective/well-tested drug to try, and for continuation of projects that did not produce anything in last 10 plus years. Such amount can produce 100s of R01 to serious investigators who could actually come up with some effective drugs.
      Also, these recommendations do not take awards like P30 and other center grants into account. Not sure why..

  89. Post doc factories versus multiple, independent investigators. This change is long overdue. Communists? Seriously? Oye….

  90. I strongly support limiting grants and I hope this plan will not be thwarted by the powerful big labs

  91. These types of changes are long overdue and many of us have been begging for such changes for a long time. We have to wait and see how well these changes will be implemented without many loopholes.

    Hopefully, this will minimize the culture of creating Masters and Slaves within scientific community.

  92. These comments are off on tangents. The 1st step is to acknowledge that there is a problem. Any scientist not living in a cave, and as outlined by Dr Lauer, would look around + notice the substantial and ongoing loss of productive mid-career scientists, with increased proportion of NIH funds going to fewer scientists. To repeat these are/ were productive, mid-career scientists. The diagnosis is easy- as with health care the solution is problematic. The NIH is on the right track but my solutions would be somewhat different as follows:
    1) increase modular R01 to $350,000 so that a single R01 really can support the proposed research.
    2) Limit PI total salary support to 50%- this frees up more money for the research, reduces PI financial incentives for multi-R01s, and stabilizes workforce by forcing university buy-in (hard money).
    3) Ask universities for full reporting (transparency) on indirect funds and reduce when they are not going directly to support of research.
    4) Lastly- change review criteria to include evaluation of research proposal in the context of PI’s NIH-supported research portfolio. Those that are strong but essentially duplicates or incremental to existing funding should receive lower score. Along these lines NIH could consider aggregate and individual criterion scores (Significance, Innovation) in making decisions about funding.
    These changes would stabilize and improve the health of what is now a seriously threatened US scientific community. The time for change is NOW.

  93. NIH funding is perhaps the most transparent and fair system in place today. All our discussion is only aimed at fining inefficiencies and observed problems in the system. The fact that NIH does not interfere with projects, once funded, allows considerable flexibility and progress.

    However, any system that remains unchanged for a while tends to become gamed and develops weaknesses. The major weakness here is that extremely well funded scientists fail to achieve proportionately higher impact work. One may examine this from the viewpoint of an investor. If you had a productive person with a productive project, will you dilute the founders leadership skills by awarding for additional irrelevant startups or take risks with projects from other promising founders?

    A second major weakness is that established scientists are being lost, creating a gap in the chain of creativity, as funds were sequestered for early stage investigators by one end of the model and well connected investigators were getting large program projects and other large grants that are often reviewed individually to sequester funds at the high end. Institutions were trying to grow without much direct commitment, as a third major weight that is slowing the system.

    A third weakness is that the system cloisters creative young people during their peak risk taking age in ivory towers only to hold a large fraction of them in underpaid insecure positions. This prevents the field from growing like software, which boasts the development of most of the top new private industries such as Alphabet, Amazon, Microsoft, Apple and Oracle.

    Given the need to make biotechnology become as successful as the software industry, it is useful to examine the most successful models of growth and encourage them while reducing or eliminating questionable regulations that hold the industry back.

  94. NIH funding is perhaps the most transparent and fair system in place today. All our discussion is only aimed at fining inefficiencies and observed problems in the system. The fact that NIH does not interfere with projects, once funded, allows considerable flexibility and progress.

    Given the need to make biotechnology as successful as the software industry, it is useful to examine the most successful models of growth and encourage them while reducing or eliminating questionable regulations that hold the industry back.

  95. It appears that the entire initiative to put caps on funding is predicated on a single plot, a log-log representation of a productivity index as a function of funding level and the observation that it levels off.

    First, note that initial presentation of Dr. Lauer also contains a similar plot that does not level off when considered beyond 6 years. This shows that productivity, as defined by NIH, may indeed track funding.

    Second, note that apparent productivity gap is only apparent in the log-log representation, showing that after the 1 R01 productivity, as defined by NIH, does not increase with the same exponential rate. Nevertheless, two R01 funding level has the triple the productivity, as defined by NIH, compared to a single R01 and keeps going up.

    Finally, the critical issue here is whether the productivity index is appropriate before even giving any of these plots a second thought. Do the review scores predict productivity, as defined by NIH? Does the productivity index predict the big leaps in science or just paper output?

    NIH should put more effort into improving the review process, and refocus on MERIT, before considering direct and drastic interference in funding levels.

  96. While I can see how this GSI points system might benefit the greater population of researchers, personally this would not be good for my research program. Specifically, I take issue with the proposed points system (rather than direct-cost $$), in particular the way it would assign points to multiple-PI grants. The choice to make a multi-PI grant count 6 points when a single-PI grant is only 7, will surely kill collaboration?

    In addition to holding an R01 as single PI (modular budget), I have a multi-PI grant with one other investigator, for which we were awarded a combined annual budget of just under $170k to split between the labs. Thus, on paper it looks like I have another R01, but in reality this multi-PI grant brings just $85k to my lab per year. On the points scale these 13 points would count virtually the same as a “14 pointer” with two single-PI non-modular $400k grants. Do I need to spell out that $800k vs. $335k (250 + 85) is a bit different in terms of scientific scope?

    I can see that a cap of 3 single PI R01s is a perfectly reasonable number, but making multi-PI grant count almost as many points, is just wrong. Why would anyone want to use up 6 of their available 21 point “quota” on a multi-PI proposal, when they could just have it all to themselves for 1 more point? Instead, it would be fairer to make the multi-PI grants somewhere in the half-way range (say 4 points). Or, you know, given the humongous variation in the actual funded budgets of all R01s, you might actually consider basing this points system on something like direct costs?

    Regarding various comments here about big/rich labs and people needing to retire, a bit of a reality check is needed about what one can actually buy with an R01 these days. Many (most?) Universities run their faculty on “soft” money, such that they’re required to provide 60-100% of salary from grants. AAUP salary scales for faculty are here (https://www.aaup.org/our-work/research/annual-report-economic-status-profession), and let’s say a mid career Associate Professor is pulling down $110k. and has to cover 70%. With a 30% fringe benefit rate that’s $143k, so 70% would be $100k. Throw in a post-doc ($47k plus 25% fringe) and a tech’ ($35k plus 30% fringe) and a grad’ student ($30k all in) and you’re at $235k out of a @50k modular budget just for salaries alone. That leaves a grand total of $15k a year to do everything else (animals, chemicals, antibodies, supplies, publication fees, travel, equipment service costs). Bottom line – a single R01 gets you a very modest 3 person lab (1 each of post-doc’, tech, student) and not much to spare. A 21 point, 3 x modular R01 situation, assuming no sharing with other PIs, might get you a max’ 8 person lab’ (2 PDFs, 1 tech, 1 staff-sci, 4 grad students). The huge labs with 10+ students and just as many post-docs, are doing it with endowment money, training grants, running the big equipment like a core so they get their staff scientists paid out of core budgets, etc. Science is expensive (my mouse budget alone is $25k a year), and it is simply wrong to cap funding in a manner that ignores the key determinant of productivity – how many dollars a PI has available.

    1. Good points. This is one reason for limiting salary coverage to 50%. Also, it will make sense to set the point system by actual federal dollars to the PI/lab rather than arbitrarily assigning numbers to grant types.

  97. I will join the voices noting that it will be terribly disappointing if this scheme does not account for MultiPI (and even Co-I) division of grants. This will be a significant motivation to keep all the resources within one group, thereby dismantling all of the progress that has been made to encourage collaborative projects.

    Given that there is already disparity of funding of women on the larger mechanism and for African-American PIs for RPGs, what will you be doing to ensure the “review” process for exceptions to the 21 points does not further limit opportunities for women and minorities?

  98. This proposal moves in the right direction, but it needs some recalibration to ensure it is equitable. Firstly, the points system awards too many points for multi-PI grants. Typically a PI has much less salary and research support on an MPI application than a single PI grant, so 6 points instead of 7 seems excessive. This will discourage MPI grants, and instead encourage would-be MPI applications to evolve into single PI applications that include co-investigators with very substantial roles, which defeats the spirit of the proposed change.
    Secondly, T32 grants should carry no points, as the PI cannot derive salary or research support from such grants. If points must be assigned, then there should be a requirement that the institution contributes salary support to cover the effort spent on administering T32s, which would be a positive change even under the current system.

    There should also be consideration for the amount of salary support an individual PI needs to acquire from grants. Some institutions require 100% grant support, so mid-career PIs will need at least 2 or even 3 grants to cover this amount. Other PIs may have 100% hard money support, so their financial requirements for salary are less. This should be factored into the system as some type of sliding scale.

  99. As someone who had reviewed the NIH intramural program I can testify that there is a lot of waste at the NIH , especially when it comes to non productive and mediocre mid level investigators , who would not have survived in the extramural world. Why not go after them? Finally , you proposed system of outing numbers on grants given to individual investigators is draconian and totally unfair. What about HHMI investigators who already have a million or two from HHMI . Why should they be allowed to obtain a single or even two R01 grants under your new proposal? I hope that you post this this response and will let the public discuss it. It is not up to the NIH administration to decide what to do with our tax dollars.

  100. We, members of the scientific community , would like to know how the NIH arrived to the conclusion that research dollars do not equal scientific productivity. First , please be totally transparent and tell us how you define scientific productivity. Is a paper that describes the outcome of a six month investigation by a single investigator equal to a paper that sums up the work of a large team that had worked together for three years? Second, where is the evidence that very senior investigators whose laboratories have been making major research breakthroughs , developing new technologies and experimental systems and provide training to scores of postdoc a and graduate students , do not provide a bigger or at least equal return on the public investment in biomedical research. Third, if there are well funded investigators out there with minimal scientific productivity , whose fault is it for continuing to fund them? It’s the NIH fault! But this does not justify a witch hunt and a crusade against so called well funded senior investigators who take scarce research dollars away form young investigators , who in many cases have trained in those big , so called unproductive labs. In the same token , we need to be informed in a totally transparent manner , about the research productivity of the intramural program

  101. After reading through comments, it is unclear why total $$ for each PI won’t be considered in this case. Based on the current funding levels, the direct costs for three R01s are around $600K. If the PI was in a soft-money institute, it only barely covers her/his salary plus three or four postdocs’ and supplies. The situation would be even worse if the lab uses animal models. Is this type of lab considered as a “big lab”. why do we have to eliminate this type and size of labs. If yes, why NIH leadership did not plan to make a changes for those “huge and gigantic” labs, which always have more than $1M annual direct costs, including our NIH director’s lab!!!!

  102. The devil is in the details.
    1. If an R01 is multi-PI, each PI should only be assigned points for it relative to their contribution – not 6 points each, for example, if a sole-PI R01 is 7. Otherwise you will squelch collaborative work, and the important cross-cutting findings it provides. For example, for a MPI R01 with 2 PIs with equal effort, each PI should be assigned 3.5 points. If you need round numbers and an R01 is 7, one PI (the contact PI?) could get 4 points, the other PI 3.
    2. Non-modular R01s should receive different points-values from modular. I’ve seen non-modular R01s with annual direct costs that are double those of a modular grant. It would be unfair to assign the former the same 7 points as the latter, given that the driver for this new policy is to spread grant-money to more labs.
    3. The entire two year budget of an R21 is only one module more than a single-year’s budget of 10-module R01. An R21 should not be assigned more than 2 points, if an R01 is 7.

  103. I think there’s a simpler solution, and one that is already implemented by another agency. Include the cost of overhead in the budget, as NSF does, rather than added on afterwards. This will have 3 direct, meritorious effects:

    1. Institutions will no longer have as large an incentive to negotiate indirects upwards, since doing so will decrease the budget available to their faculty. Thus the constant pressure on indirect costs (including new building) will diminish.

    2. High overhead institutions (LBL, Broad, Salk) are typically stand-alone centers with entirely soft-money populations. These are the places with the greatest pressure to achieve 3 or 4 R01s, because of the salary that must come out of the grants. By making these places less attractive, more researchers will choose to affiliate with sites with lower indirect costs, saving the NIH money.

    3. This system is far easier to administrate than a complex points-based assignment that is subject to all kinds of gaming and complaints and exceptions, as seen above.

    Of course this would be paired with an increase in the standard R01 budget to, say, $375K per year. Then at a 50% indirect rate institute, the researcher would receive $250K per year, but at a 70% indirect rate institute, a researcher would receive $220K/yr.

    Best of all, it’s hard to argue with the logic; there’s no reason the government should be paying more for science that happens from an R01 given to the Broad Institute than for an identical R01 that goes to Indiana University.

  104. Many comments are about postdocs being slaves to their PIs. This is very convoluted logic and will not change with this new system. Post docs have to train somewhere. It doesn’t really matter if they train in a large, well-funded lab or a small lab – they still spend 3 or more years overworked and underpaid. This whole issue is distinct from the issue of PI productivity and should be considered separately.

  105. For laboratories at institutions that do not support salary or ongoing operations, the limit to three major grants will be quite severe. It will lead to downsizing of labs that produced leading edge science and will stifle progress by prohibiting access to new and costly technology. I consider myself “mid-career” and project that we will have to downsize considerably over the coming years. Importantly, the currently proposed system will dis-incentivize collaboration in multi-PI teams and PPGs (I just canceled a planned application).

    If the concern is diminishing returns per tax dollar, an adequate solution would steer resources to where they make a difference, for instance with a system that scrutinizes (and rewards) output per dollar invested in the last 5 years. This could be done individually for each applicant during the review process.

  106. I like the overall idea but one serious flaw in the implementation is the treatment of multi-PI grants. A solo R01 is 7 pts but a multi-PI R01 is 6 pts. This means that those who rely on collaborations, getting a 1/2-R01 or a 1/3-R01 for each collaboration, are screwed. In neuroscience, at least, collaborations are pretty much the only way that theorists can get NIH grants these days. There’s got to be some accounting for the money received in a multi-PI grant to decide if it is effectively a full grant or 1/2 a grant or less, with points assigned accordingly. For example, an R01 that gives the PI less than some threshold — $125K/yr or $150K/yr direct costs — should be assigned 3.5 points instead of 6 or 7. And similarly for other forms of multi-PI grants.

  107. Although I have very little to add to the many good comments above, commenting here is more of signing a petition to stop the proposed point system or at least to think through it very carefully prior to arbitrarily assigning points. After seeing this point system, my first thought was to give up on our MPI grant (where my contribution is crucial) and my second thought was to stop working on P01 with three other junior/mid-carrer PIs. This new approach will kill the collaborative research! I am not saying that at some point the correlation between invested $$$ and productivity stops being linear, but 3 average R01s (roughly equal to $600-700K direct costs) cannot be that mark!

  108. This is a potentially helpful way to free up funding for more researchers, but as currently configured, the points system discourages collaborative research and community contributions. Researchers now have a huge disincentive to contribute to collaborative projects with small budgets for each contributor, because the points would take away from single lab projects that would provide more support for the PI and the work of one lab. It would provide a disincentive to serve as PI of a training grant, which costs a couple of points but provides no support for a lab’s research. It would make it a disincentive to serve as a PI of a service or training core because the work would be done for others and the points would prevent being PI on a research project. All of these initiatives provide very little money to the PI (a fraction of salary) but would count against applying for money to run the PI’s lab. These unintended consequences will have an immediate effect on collaborative science and training and lead to an even more self-centered scientfiic workforce. I urge you to reconfigure the points system to make sure there is not a profound disincentive to contributions to training and scientific collaboration.

  109. I am one of the “6 percent” who works at a “soft money” institution where investigators are required to pay 80% of their salaries from grants. I recognize that the intent of the proposed cap is laudable. However, I do not understand how assigning points to recipients training awards (T32) is helpful to the NIH mission. In essence, this would make senior people with extensive and successful training records out of the running to renew their T32s if they already exceed the proposed GSI of 21. The more junior investigators who are competing for their second R01 (an absolute necessity in soft money institutions) have neither the track record nor the administrative time to credibly lead a training program.

  110. Generally, I agree with the premise that good science should be supported. Support decisions should be blinded from any concerns other than ‘is it the best science for the funding purposes and money available from the granting organization’. If multiple instances of good science come from a single laboratory, and that laboratory continues to demonstrate that it is an adept steward of those funds by proposing a ‘fundable’ follow-up proposal, then it would be foolish to discontinue that project simply because there is an artificial and arbitrary cap on the number of NIH grants an individual may have. That is counterproductive.

    Of course, there are instances where funding decisions are designed to collectively think about the long-term propagation of science. Hence, the decision to provide some form of preferred status to applications from new investigators or by divvying up NIH funds into different institutes with distinct purposes. The essential question is what benefits arise from a cap? The only potential benefit is whether it distributes funding across more laboratories, leading to a bigger source from which new ideas may spring. However, that only works if the bigger pool of investigators is sustainable, which comes only from increased funding since, as noted below, insufficient funds distributed across a larger pool, only causes more to drown and thereby shrink the pool size.

    There certainly are concerns about a cap on NIH grants related to laboratories operating on 100% soft money (for which 2-3 NIH grants is insufficient unless each of those grants is large). We’d have to address the soft money concern before implementation.

    There also are concerns about whether a cap impedes following new and unexpected/exciting leads since, if a laboratory is already capped out, that means the laboratory must shutter another equally outstanding program to follow that lead. So, the lead is not followed. Even worse, the lead is not followed and another program within the laboratory is not renewed, which leads to the dissolution of the collective expertise of what had been a functioning team. Keep in mind that, when an effective project team is shut down, it takes years to develop that again and most likely never happens. If there is a cap, there has to be some understanding then of the best way of keeping projects going when the ceiling is reached. Some ‘new investigator’ provision that helps transition that new project to someone else is laudable and even encouraged. But it is not always the best way to continue a project when the existing PI had been the actual driver of the progress.

    In short, I do not see that the cap, by itself, offers any solutions to today’s current funding problems. I can even envisage ways it which it may impede scientific discovery. I recommend against the implementation of such a cap.

  111. Great discussion. As a prior young PI who left science due to the funding crisis, I would like to add to the comments regarding capping or reducing indirect costs. One way to do this is to analyze and reduce excess consumption of laboratory buildings and processes. Labs are among the highest energy consumers on a university campus, and oftentimes, these levels exceed what is necessary to safely conduct research. Identifying inefficiencies in the system and reducing these could reduce the overhead costs of research, thereby increasing the bucket of funds available for direct costs and to more investigators. For example, we could be better at sharing equipment, lab space utilization, reducing hazardous waste generation, and reducing energy and water consumption.

  112. It is understandable that the NIH would consider a new policy such as Grant Support Index (GSI) in order to ensure equity across the grant portfolio. But at the same time the devil is in the details. While we all want to ensure that young investigators are not only able to receive their first NIH grant, but their second as well, we should balance that goal with the need to ensure the NIH funds the best most impactful science. We must also ensure that we don’t restrict innovation.
    If an investigator is producing high quality science that has the potential to positively impact human health they should not be blocked from receiving additional support. At the same time if a researcher is collaborating with another successful investigator, they should not be restricted from further support. The NIH has stated the great need for interdisciplinary research, so it is important that the GSI policy does not negatively impact investigators ability to collaborate. If researchers are concerned that participation on a muli-investigator grant will negatively impact funding for their lab, innovation will be stifled.
    As the NIH finalizes the GSI policy, I urge that a nuanced approach be taken that enables high quality interdisciplinary science, while ensuring that the next generation of scientists can get their careers off the ground. While $34 billion is not enough funding to support the full biomedical research enterprise, the NIH budget is large enough to be able to develop a balance GSI policy that achieves multiple goals.

  113. NIH needs to focus on cut funding to big center grant, intra-mural funding as well as small business grants to save money instead of cutting R01s. Those grants have more waste per capital and tends to have heavy administrative support. It is not well justified to limit resources on merit-based basic research funding that is already highly competitive. The limit is going to be counter-productive and bad for science.

  114. I am an assistant professor with a lab less than 4 years old. I am not at a famous or IV league school. Talented assistant professors do not need this point system to succeed. We can get RO1s by writing excellent grants, and by collaborating with the right people in the field. I am afraid that those collaborations would be difficult to come by for young investigators who need them, should this point system go into effect. With “only” 3 RO1s, I would have to cut down significantly on my in vivo work, and still not have sufficient staff to do the work because my salary contributions would have to be increased dramatically to meet my schools expectations. That way, my productivity would suffer. I would probably not be able to collaborate nearly as much as I used to, would not consider writing project grants or T32s, and would try to convert some current grants (especially MPI grants) into Co-I grants to make room for what I consider more creative, interesting science I can lead. I would expect senior investigators to not want to collaborate as much any more either. I would not submit MPI grants with people outside my field any more for what could potentially be important collaborative work.

    Some well-funded labs are unproductive, some are very productive. Some assistant professors are talented and in the job well, some do not. Point systems don’t help. Grant renewals should be judge by productivity, as in numbers of papers per proposed staff or $$, and impact of papers (journal, citations). We need to have a fair review process that emphasizes productivity. To support new investigators / early investigators, we should increase the number of grant opportunities exclusive to them for them to have the opportunity to demonstrate their productivity, but not do that by having a point system. If some RFAs are made early / new investigator only, then there are fewer RFAs to compete for more senior investigators, however, it is at least still a competition.

  115. Wish you guys had thought of this idea 20 years ago – maybe one or more of my “near-miss” grant applications would have been funded. There is already a problem (from my friends serving on NIH study sections) of the Catch-22 that unless you have a track record (or someone in the section who decides to advocate for your study), you get scored poorly. Even when I applied for R21 grants that specifically stated you did not need preliminary data, the grant would get cited for lack of preliminary data. If I submitted a grant with preliminary data, AND highly respected and funded collaborators from one of the top institutions in the country, well then there was no evidence of funded prior collaboration. The consistent pattern – seen both by myself and by friends across the country – is that you only get funding if you have previously succeeded in getting funding. The exception, of course, are K awards – but if you start down the research path too late for a K – well, you are out of luck. This mindset led directly to a small number of researchers receiving a disproportionately large percentage of the grants and grant money. I won’t even discuss those situation where grants are rejected by the study section, then one of the section members steals the idea. Illegal, unethical, and yes, it happens.

    Limiting the number of awards per researcher is a start, but a major reform of the way peer review is conducted is also critically necessary. It is too late for me, but it is not too late to prevent the loss of another entire generation of potentially brilliant ideas that will not see the light of day due to lack of funding.

  116. It has been said before. But any good system will incentivise not discourage collaboration. I have many grants. And relatively little money. The new system would financially disincentivize me from starting any more collaborations. Leaving the points for the multiPI grants the way they are would basically abolish the multiPI grants for anyone even moderately successful.

    Worse, it creates an incentive for the top people to maximize the cost of their small number of grants and prioritize expensive science that brings overhead over cheap high impact research.

    Also, as much of the proposal seems driven by the graph with the nonlinearity: it would be nice to provide some causal evidence instead of confounded correlations.

  117. I have heard from multiple sources that institutes like NIGMS have already instituted caps previously and also recently created the R35 mechanism which was supposed to address many of the issues brought up as reasoning for instituting a cap. Before putting in any new policy, can analysis be done to determine whether NIGMS or other institutes that have already put caps in place fund a larger percentage of unique PIs relative to institutes that do not have caps? Also, did the R35 mechanism increase the number of unique PIs or change the number of early or mid-career scientists funded relative to the older scientists? It seems to me that NIH may already have some data to either support caps or not and this would be useful to the discussion and also policy making.

    It also seems to me that without an accompanying policy that directly gives some advantage to mid-career scientists (like an ESI type advantage) or directly to those with only 1 R01 who would lose their lab, the net effect of this type of change would do little to address the problems set out.

  118. Over my 35 years of research experience in academic institutions, my observations of how labs of varying sizes (10-20 members ‘large lab’, 5-9 ‘mid-size lab’, ‘1-4’ ‘small lab’) operate have led me to conclude that the productivity and quality of work are not necessarily proportional to the amount of grant money/number of lab members. There are several common problems in large labs: (a) the PI usually does not have enough time (or even insight) to sufficiently monitor potential problems in individual projects, (b) those that are at the bottom of the food chain (e.g. junior students) do not have sufficient guidance from the PI, (c) there may be mismatches or errors in experimental design/execution/data analysis/interpretation when experiments are done by different lab members without sufficient communications among them and with the PI.
    Here is an example I witnessed in my neighbor lab (large size): a project was designed to study how a small molecule interacts with a voltage-gated channel. The experimental data support interactions in the closed state of the channel, yet the molecular dynamics simulations were carried out using the open state of the channel. The paper was published in a high-impact journal. The body of the paper reported interactions in the closed state, and the Supplemental Materials report results of molecular dynamics simulations without specifying which state of the channel was used in the computations. I wonder: what type of message those postdocs and students in the lab get from such an experience?
    Limiting the grant money each PI can have is a good way to provide more chances for small-mid size labs to thrive. I believe this is the smart way for NIH to invest its resources for the continued success of biomedical research enterprise in US.

  119. An Alternative to the Proposed GSI : Two major reasons cited for use of the proposed GSI as a means to curb the numbers of grants allocated to an individual Principal Investigator (PI) were articulated:

    1. Scientific productivity tracks weakly with funding
    2. A need to foster younger and mid-career investigators

    It is dubious that the proposed GSI point system would effectively address the goal of funding the best biomedical science and ensuring an optimal future biomedical research work force. Concerns with GSI have been articulated by many others on this blog, in particular concerns about penalizing training, inconsistencies with current NIH percent effort standards, and no systematic approach to foster younger investigators.

    We propose an alternative approach that will not require a great deal of re-engineering and will more effectively allow for production of the “best science” and an ongoing effective training mechanism.

    1. Linking scientific productivity to funding for Highly-funded-PIs: Investigators could apply for as many grants as they want, adhering to NIH policies with regard to percent effort required for each grant. However, for those PIs whose NIH research funds (excluding any funds from training grants) exceed a certain amount per year (for example $1million/year, termed hereby Highly-funded-PI), an objective measurement such as Relative Citation Ratio (RCR) (https://nexus.od.nih.gov/all/2016/10/21/applying-the-relative-citation-ratio-as-a-measure-of-grant-productivity/) could be used by individual Institutes to limit the fraction of Highly-funded-PIs able to obtain additional funding.

    In this scenario, the ability of a Highly-funded-PI to obtain a grant over and above the $1 million per year limit (in the example given) would be determined by their RCRs relative to those of the other Highly-funded-PIs in their cohort. An individual Institute, for example, could choose to fund only 50% of the Highly-funded-PIs, who would be chosen as individuals whose RCRs were in the top 50% of the overall Highly-funded-PIs in the same funding cycle.

    An individual Institute might decide to draw different lines for investigators with different amounts of funding. For example, for Highly-funded-PIs with direct support between $1 million to $1.5 million, an Institute might fund the upper 50% based on RCR ranking, but for Highly-funded-PIs with direct costs between $1.5 million to $2 million, the Institute might decide to fund only the upper 25% based on RCR ranking.

    2. Fostering younger and mid-career investigators: Funds saved under the paradigm discussed above could be used to guarantee that ESIs (Early Stage Investigators) and mid-career investigators (MCIs) have a fair opportunity to obtain funding. Currently, some NIH institutes fund ESIs at percentile rankings that are more generous than those used for more established investigators. For example, the NHLBI funds RO1’s for ESIs at the 25th percentile, whereas for more established investigators, RO1s are funded at the 15th percentile. Although this system is designed to promote funding of young investigators, it does not ensure that the playing field is leveled.

    One way to ensure a level playing field for ESIs/MCIs would be to calculate the fraction of ESIs/MCIs who are being funded with this system, relative to the fraction of ESIs/MCIs that were funded during a period when overall NIH funding levels were greater than at the current time. These calculations should also take into account the fractional representation of each class of PIs at each time period (plentiful versus lean).

    If similar fractions are currently being funded, the playing field has been leveled for ESIs/MCIs in this relatively lean period. If, however, a smaller fraction of ESIs/MCIs is being funded than has been historically funded by NIH during a more plentiful budget period, the percentile advantage for the ESI/MCI would have to be increased.

  120. The proposed point system will significantly impact collaborative research programs and undermine the collaborative research agenda that NIH has been trying to build for the past many years. Multidisciplinary collaborative ventures like P grants involve many investigators and have tight budgets and as a result none of the components of these grants are really well funded. If the current point system is implemented most investigators will not consider it worth their while to collaborate with others, or participate in collaborative ventures.

  121. The new 21 point system is misguided for the following reasons: 1) it would limit collaboration on multi-PI grants as being a mPI is a liability for points – costs 6 points for the contact PI and 5 for the second PI – total of 11 for a mPI R01 vs 7 for a single PI R01. I would immediately turn over my mPI grants to the other PI and not do any new ones. 2) The points for a P grant (eg PO1, center grants) are way too high – the program director gets severely penalized on points, for very little additional money over them doing a regular R01. In general, I would not do any more collaborative grants. 3) The points for small grants are too high. 4) There will a move to increase the $ value for each grant submitted to get the most for each set of points. Not sure that will help spread the wealth. On a slightly different topic but related to the need for a system where merit, not wealth determines grant success, expensive projects (eg imaging, mouse work) require an R01s worth of funds just to fund the necessary costs associated with microscope/magnet work, or mouse costs. These could be dealt with as grant “add ons” or supplements reviewed at the time of the parent R01 review. Otherwise it becomes too expensive to do this sort of work. The same goes for working in an expensive location. I think the points policy is misguded in general – productive and supportive well funded labs shouldn’t be limited – they should be enabled. In addition to it’s contribution to science, a well funded lab supports many junior personnel who need the support of their mentor’s funds to get to the stage where they can compete for their own funding. Another system than supports junior/less experienced researchers is certainly welcome, but not at the expense of more senior/experienced/successful PIs. The proposed policy needs seriously rethinking.

  122. The point system does not seem to encourage team science and community-wide research projects or initiatives. I have a cross-disciplinary research lab with team members from diverse training backgrounds. We collaborate with multiple other labs and are reasonably productive (with an average of 5 publications per team member). If there is only one point difference between Single PI (7) and MPI (6), we cannot sustain the lab as the lab only gets its share of the grant budget in most of those team science projects (half, one third, or even smaller). I believe my lab provides an excellent training environment for team science researchers as almost all of my postdoctoral trainees end up to be a successful investigator.

  123. This is one of the worst ideas that NIH has come up with in a long time. The concept of “spreading the wealth” has been tried by many agencies especially in Europe and in Canada (I am most familiar with the latter). All this mechanism ends up doing is supporting mediocre research. If NIH really wants to support new investigators and mid-level investigator renewals, then set up specific programs for this. Let the “highly” funded researchers compete among themselves. Some labs are WAY over-funded, but there are also many more that are highly productive, and have spent decades to attract, train, and retain the personnel that contribute to the success of those labs. What will become of those highly qualified (and in most cases, well-paid) researchers? Are we just to discard them in favor of some new initiative? That would be beyond unacceptable.

    PS. I am now retired from active research after 40 years, so I have no vested interest in the success or failure of this initiative.

  124. I hope a second set of comments will not be unwelcome

    I personally agree with the notion that some investigators can become professionally overcommitted to the point where additional grant funding may yield diminishing returns, no matter how talented the PI or how good the idea the grant is based on. I also agree that NIH is right to try to come up with a formula to determine where the points of overcommitment are and impose limits on additional funding. Reinvestment of the funds freed up by application of this formula into the careers of underfunded investigators, especially, young investigators is a good idea.

    However. I respectfully offer my impression that the rollout of this program appears to be rushed and unvetted. The GSI formula remains unavailable. All we know is that it may resemble the “Research Commitment Index” (RCI) published at:

    https://nexus.od.nih.gov/all/2017/01/26/research-commitment-index-a-new-tool-for-describing-grant-support/

    The RCI has obvious flaws. It can be gamed and it also discourages collaborative research. (See my May 18, 11:10 AM comments posted at this same “Open Mike” blog). I also think that there are likely to be numerous unforeseen consequences to universal application of the GSI to all extramural grant PIs. Review of the final GSI formula will be very important when it is actually drafted. Indeed, why not do beta testing to de-bug this program before deploying it?

    Finally, what is the formula for the redistributions of funds that will be freed up by application of the GSI? Is there, in fact, a plan? Is it a good plan? Have all relevant authorities signed off on the plan?

    To summarize, I am concerned that the soon-to-be-imposed GSI-based program has not yet been thoroughly vetted or beta-tested and that there is insufficient time for it be so prior to a planned deployment starting this fall.

    Best wishes to all!

  125. I completely agree with the commenters that have expressed concern over MPI R01 grants receiving 6 points, compared to single PI R01 grants receiving 7 points. This will lead to fewer MPI grants, which would be a real shame, since this has been one of the best NIH implementations recently, which really encourages collaborations. The MPI R01 grants should be split as 3.5 points for a 2 MPI grant or 2.3 points for a 3 MPI grant. The same should apply for a MPI R21 grant, which also encourages collaboration on high risk projects.

  126. Disclosure: I am one of the PIs who woudl be well over the RCI/GSI cap of 21. I am also a firm believer in faculty development and in mentoring them to establish independent careers. As a former department chair, center director, and now institute director. I view one of the highlights of my career as supporting career development for junior faculty. Many of those who I have trained, hired, or simply mentored, have gone on to establish strong NIH funded programs. So I am highly supportive of efforts to insure that younger PIs can compete on a level playing field for funding, that being said I am not sure that the RCI/GSI system is the best or even the most facile way to do this. I will first express my concerns over the RCI point system as proposed (though I know this is simply a proposal and not necessarily the final roll out). I will then suggest several alternatives that I believe could accomplish the same goal without imposing rigid criteria that will artificially create winners and losers.
    Multiple concerns over RCI/GSI.
    Let’s compare a direct costs $1M P50 with a $1M single PI R01, the later assigns 7 RCI points to one investigator. The former assigns 11 to the overall PI, and 6 to the director of each component. So with 7-8 components one ends up total of 53-59 RCI points. This makes no sense!
    For those like myself who run a P50 research center (largely an administrative and vision level function) my wet-bench lab actually gets ~$25,000 in support from this grant. Yet this grant supports over 25 faculty including junior faculty and in our case three junior faculty who are underrepresented minorities. Further in order to make this multi-institution grant run I provide support from my institution and benefactors of ~$500K. But in the current scoring system I will be forced to chose between successfully running this grant and successfully running my lab. I could not support my wet-bench laboratory on a single R01.

    Few will want to participate in big grants like this unless that is the only mechanisms by which they could be funded. I would suggest that there must be a link between RCI, the overall funding level, and total faculty FTE/key personnel effort and that the RCI points are distributed for larger grants in a very different fashion. So if a P50 is ~3X the funding of an average R01 then there are 21 RCI points distributed across all components and the overall lead. This could be scaled to varying degrees, but would still encourage collaborative science.

    The lack of scaling RCI to funding also a big problem. There needs to be some scaling relative to dollars. Grants of the same activity code vary hugely in their funding and RCI needs to be scaled to reflect this.

    Finally for those like myself who may have had R01s that have been funded for 2 or 3 cycles and who are “over the RCI limit” are you really going to say you have been successful but you can’t try and renew your grant simply because your over the “cap”. I happen to work in one of the areas where NIH funding has dramatically increased and on a disease, Alzheimer’s, where there is huge unmet medical need. I think we are doing some of our best and most translation work that could in the future benefit patients. I admit I have some hubris here, like most scientists, and I do not think anyone else can take over and drive the science. I have spun multiple projects already to several junior faculty that they have been independently funded. I do not have anyone else to take over. Does somebody really want to defend this policy to a taxpayer when a disgruntled scientists, says I think I have a cure for X that is even partially legitimate, but that the scientist could not get more funding because the had to many RCI points?
    Further, it is large labs like my own that support senior staff scientists. The jobs of these individuals will be endangered by this policy as they can only be supported by there FTE being distributed across multiple grants. This might create opportunists for more junior scientists, but if I am a staff scientist with a reasonable salary, I woudl be worried.

    Some Alternative Strategies.

    1. Peer-review and funding linked to rank of investigator. Some of the rationale for the RCI/GSI was to provide more opportunity for mid-career investigators. Why not proportion funding to faculty at each rank (assistant, associate, and full)? Then decide how to reapportion the current funding to support PIs at each rank. Further separate each rank into separate study sections so you are funding the best science at each rank. This seems to have worked reasonably well for young investigators (not reviewed in separate study section but as a group). So why not go the distance here? This might mean for Professors paylines are quite a bit lower than at each rank before, but don’t penalize somebody for success. If someone can compete still let them. Ultimately this will insure that the best science at each rank is funded and senior PIs will not have an unfair advantage but have to work even harder to insure they remain competitive. This should insure that there is a good work force at each rank.

    2. Increasingly limit the size of R01s that are awarded when a PI has more than one. I think this would actually involve needing to move the size of the first R01 up. But one could imagine a system whereby the first R01 can be funded up to X dollars the next up to .8X and the next at .6X. There are typically synergies when one has more than one R01 and that can be reflected by reduced costs for the next one awarded. This means one has to work incrementally harder to get more grants but still leaves funding as a meritocracy. albeit one with imperfect peer review.

  127. I have sympathy for the goal of broadening support across the scientific community. But this specific proposal will have a terrible negative impact on labs like mine that engage in a lot of collaborations that result in multiple awards that individually provide only a relatively small sum of support. Implementation in the current form would thereby strongly discourage collaboration. We need a different way to solve the problem. For example, if a points system is imposed, it should only apply after a lab reaches a three RO1 ($750K direct) level of funding.

  128. One of the changes that NIH could consider that could easily result in more available grant funds for more labs is changing the carryover model. Labs that spend with a focus on efficiency (buying cheapest materials, used equipment, finding ways to do experiments at lower cost, etc) can have funds left over at the end of a project, sometimes substantial amounts. However, because NIH penalizes large carryover and makes no-cost extensions more than one year challenging to obtain, many labs feel pressure to spend their money within the original budget period rather than conserve it, and therefore are not as efficient as they could be. We all know that paper towels from major scientific distributors are 2x more expensive then at the local big box store but many labs buy from the scientific distributors because they are convenient – there is room for efficiency. I have seen this model in many labs, large and small. It is the same inefficient model with university startup dollars which need to be spent in a short fixed term. If NIH allowed automatic carryover of funds without scrutiny and allowed NCEs of 2-3 years automatically – you could universally guarantee that labs would find how to conserve every dollar they could to make that money last as long as they could, especially in today’s funding climate. Many investigators would stretch 2 year R21s to 4 years and 4 year R01s to 6 years. Who benefits? The investigators benefit by having less stress to apply for grants as often, smaller labs benefit by more available grant dollars because investigators are applying for grants less often, and NIH and society benefit because this model encourages efficiency with spending.

  129. As a young PI who trained in a large well funded laboratory, I feel I have to voice my concern over the GSI. I don’t believe in re-distribution of funding anymore than I believe in wealth re-distribution. The best science should be funded period, regardless of name and institution. It is important to note that generally it is the large, well-resourced labs that are the training ground for the next generation of PIs. These labs have the bandwidth to foster young careers and allow trainees to develop new ideas and move on to develop them independently. A lab with more restricted funding is less likely to let discoveries out the door because it would comprise the maintenance of the funding stream.

    Also, I echo the many previous comments that raise the issue of the single point difference between lead PI and collaborator. This is in short “punitive” and would certainly discourage collaboration. Arguably, the greatest leaps in science are achieved through cross-disciplinary efforts and integrated research.

    The GSI needs some serious rethinking!

  130. The idea of almost the same points for a MPI vs a solo R01 is subversive of collaborative science. The idea of MPIs was to encourage people of diverse expertise to combine on a focused problem, with the successful exploration would be enhanced as a result. With the pressure not only to discover, but also to satisfy academic departments of above threshold levels of lab and salary support, the incentive to share, in spite of the continued scientific benefits, gets severely challenged.

  131. As all of the large academic institutions and hospitals weigh in against this, please note that the opinions they provide are not those of junior investigators at those institutions, but rather the opinions of the very senior investigators whose funding you are proposing to limit. We junior investigators really struggle to survive at these institutions, and the GSI might help.

  132. The general idea of a funding cap is plausible. However, the current algorithm of point assignment may lead to un-intended side effects. A). For example, a PI of a R01 grant with 3 Co-PIs will likely get only 1/3 of the funding. If someone serves as a Co-PI on 3 of these multi-PI R01s, he/she will have 18 points according to the current NIH point scale, which is very close to the 21-point cap. However, he/she will essentially only have one R01, which is insufficient to support a strong research program. The point assignment formula in its current form will discourage collaborations: people will be less likely to participate in multi-PI projects and opt for single PI projects. This is in conflict with NIH’s policy to encourage collaborations and multi-PI proposals. B). People may also be less interested in R21 and in favor of R01 (4 points for R21 seems too high compared to 7 points for R01). C). NIH funded projects will often have no-cost extension. These no-cost extensions usually have very small budgets (and none are NEW funding), and thus probably should not have the same assigned points.
    Hope NIH can modify the point scale to make a positive impact.

  133. I am not sure this has already been proposed (I must admit I haven’t read all the comments), but rather than cap the number of grants per PI, I would suggest enforcing that PIs spend more effort on the grants they are meant to lead.
    I always question effort level below 20% for PIs of R01s. Less than a day a week simply isn’t enough time to properly supervise a research project.
    But really, any sort of cap is a challenge for everyone. Given the funding levels, everyone needs to submit many more grants than they expect to get. In the lucky event that you exceed your expectation. What happens, you give up a project?

  134. This is clearly a polarizing issue with heated arguments for either point of view. The decision is motivated by a great idea. Surely, it is important that Bioscience is vigorous, young, and diversified. The implementation of the GSI is perhaps animated by the hope that the talented, ambitious scientists will have a better chance to succeed. Yes, succeed. But to a point. Until they have 3 R01s, and then they hit a ceiling. The problem with this measure is not the motivation to give the young and middle-stage investigators a chance. It is about who gets punished.

    And who gets punished? The successful investigators. The thought leaders. The connectors. The mentors who create highly dynamic labs and train the next generation of scientists. I will not elaborate on comments that were already made by others on this forum about transparency as to how NIH judges so-called diminished productivity. But I’ll add a few thoughts. It’s worth considering that grants run for 4-5 years before they are subject to competitive renewal. This is a very short time frame if one embarks on an ambitious, high risk, and important project. A grant renewal is (or perhaps should be) a referendum on productivity, and if investigators are not productive, the system needs to recognize it. If indeed there are too many unproductive investigators with 3+ grants, then the NIH should review how it evaluates Investigator productivity (and conversely how it rewards productivity). The United States Biomedical research enterprise will fall behind (and in many ways it already is falling behind) if the NIH fails to recognize that some of the most consequential science cannot be funded by 3 or fewer R01s.

  135. There would be little incentive for an individual to host a training grant given this applies points to an individual without any financial support to the PI or the PI’s research effort. This will have an adverse effect on mentoring young scientists. The concept will further limit mentoring as PI’s will not be in a position to leverage their funds to develop independent careers. Each time a PI takes on a mentee, he/she sheds a portion of their program which has to be regenerated further diminishing mentoring. Overall, this will have the opposite effect by tending to diminish the workforce through mentoring

  136. Such a policy will reduce innovation. New ideas will have little opportunity to develop. Further, this is an artificial policy that will reduce the senior workforce by forcing the very mentors out of science who are needed to develop the talent in highly specialized fields. Please let the study sections do their work by funding the most meritorious science. Keep the meritocracy intact.

  137. This is reasonable with two major exceptions.
    First, it is not clear why a multi-PI grant should count as 6 points. This doesn’t make sense, as the grant is now being counted as a total of 12 points, rather than 7. Practically, it will disincentivize such applications unless multi-PI grants are funded at twice the level of individual PI grants of the same mechanism.
    Secondly, while the number of points credited against one’s total for being PI of a trainnig grant is small, it will discourage investigators who would normally carry 21 points of research awards from serving as PIs. These successful investigators are often the best positioned to direct a training program. A similar argument can be made regarding center grants, e.g. P30, depending on how much support the PI’s own research program receives from the grant. This could be assessed administratively or during peer review.

  138. My institution (hospital) provides basic researchers with basic researchers. In addition to the high research costs and salaries, PI’s at my institution are expected to get two R01 grants to get by. So, the power of money differ substantially depending on the institution, location, and nature of research, etc. I basically support funding cap, but I hope that NIH takes each PI’s specific condition into account in capping funds, but not blanketly apply a single criterium.

  139. As a young investigator, I welcome NIH’s efforts to level the playing field for us and make more grant money available to a wide spectrum of researchers; however, I believe that this should not be at the cost of good science conducted by many well-established scientists that are mentors to many and it certainly should not be based on a deficient and unfair scoring system. The proposal, as I have reviewed, seems to neglect the fact that not all R01s are created equal and I am sure it has similar deficiencies for other grant types which I might have missed.

    I would like to share a hypothetical example:

    Take and R01 with an MPI structure with 2 MPIs and a subcontract (co-I at the same or a different institute) at the level of 300K (a reasonable R01). In this scenario, each PI is getting about 100K to help out with the goals of the research proposal. The two MPIs are deducted 6 points each for their MPI status in an R01 while the subcontracting PI is deducted no points whatsoever. On the other hand, a 400K R01 can be easily held by a single PI costing the PI 7 points. So 100K can cost a PI 6 points while 400K can come at a mere 7 points. The MPIs in the first scenario are deducted about 3.5 times more points for the same dollar amount. As clear from this example, the GSI, as currently proposed, can grossly penalize MPIs of small R01s that are cost efficient.

    The main point of this example is that GSI should be based on “Direct Cost” basis rather than the current somewhat arbitrary approach. This could potentially make it fair for the established faculty and still make more dollars available for the younger faculty. A direct dollars based approach with the money itself as the optimization target would be much easier work with and be more transparent for the NIH itself.

    A hard cap regardless of the scoring system itself might still be detrimental; and perhaps creative solutions (for example, a soft cap with mandatory matching from host institutions beyond a certain limit) can be useful in advancing all our research and not sacrifice any good science that can make an impact.

    I thank you for your efforts and look forward to a reasonable and fair resolution by the NIH and for all involved parties.

  140. I think this is inherently the wrong way to go at the issue. Evaluation of science should be merit-based. Many of the most important contributions come from big labs. Getting some assessment of investigators’ track record is clearly worthwhile. But putting in place distribution requirements and giving more funding to inexperienced investigators at places without strong research track records or support will just waste what are very scarce resources, and will not result in more robust science, which after all is what the underlying goal is in the first place.

  141. How many great ideas can a great scientist have in his or her entire life? So, 2 R01s are enough to continue that greatest research project.

  142. The goal of providing more funding for new investigators and across more institutions is very important. It will be important, however, to ensure that there are not unintended consequences by discouraging senior investigators (particularly senior clinical investigators) not to participate in collaborative clinical research that provides infrastructure for junior clinical investigators (critical for K23 awards, etc) and provides mentoring and publishing opportunities for junior investigators.

    Few specific suggestions on metrics:
    Would make P01 and R01 similar point systems (as most P01 do not provide funding to just 1 lab).
    Would divide the total points for R01/P01 by number of co-PIs – otherwise, you are strongly dis-incentivizing multi-PI projects (which is one of the best initiatives to come out of NIH in past decade).
    Would consider U01/U24/U56, K24, training grants, center grants etc to be “negative” points – you should get rewarded for participating in these collaborative and mentoring efforts, rather than penalized. It is true that they take time/effort – but very little of this funding gets to the PIs’ lab.
    Could consider instituting a “pre-review” system or even a section in the grant application itself that when an “Over 21 Club” PI is proposing a new grant, that they need to justify the opportunities that this additional funding will provide for junior investigators and other laboratories. This is preferable to the exception after review, as Reviewers may be influenced by concern about the PIs GSI, and would allow only those grants to be reviewed that meet the criteria for exception.

  143. I am biased. I received my first R01 when I turned 44 years old. Three years later, my current GSI is 27. I always thought — still think — that the reason why I failed on first R01 was my insufficient knowledge and inability to convince my peer reviewers. Now with four NIH grants, I feel I have grown and survived through the difficult time. Yet, I don’t feel my capacity and creativity have saturated. I published well without R01 and now am happy to run an amazing lab with young, bright postdocs and students in diverse background.
    Here is my humble opinion:
    — The proposed limit to 3 R01’s is too low and unhealthy.
    — Instead, improve the review system by strengthening the criteria for scientific overlap.
    — Create a grant mechanism only for early investigators.

    1. I fully agree with these suggestions, but fear that we are debating a point system that has already been decided on by the NIH with little input from the investigators.

  144. In my mind, as a younger scientist, this means that although I’ve already been lucky enough to get 2 R01s funded, there will always be a limit on the extent to which I can further grow and expand my research portfolio. Like many, I submit multiple grants a year in hopes that somethings sticks and helps to keep us in business. When less exciting, less innovative science happens to get funded over the more innovative higher-risk proposals, the GSI point system will leave us less able to get the more innovative science supported. Furthermore, given the way the proposed GSI is structured, my collaborators and I have now discarded all ideas we had for multi-PI grant proposal moving forward. And, my mentors, who have always been well funded and very generous with their time and energy and resources for the development of their mentees, will be more strapped. I strongly agree with the general premise that motivated the proposed GSI, but I am just not sure it’s going to lead to the types of results intended. I think this proposal will appeal to the larger group of fresh scientists with great new ideas that need funding. For those of us who are just a little beyond that stage, i.e. young scientists who have just started to get some traction and still have more ideas to move forward, the proposal seems a bit daunting with respect to outlining a longer term career plan in academia.

  145. Fund staff scientists and core facilities, and organize a national core facility collective. This collective would spec. out and negotiate instrument and reagent costs with manufacturers, share information and standardize protocols. The first national cores should be in major research nodes like Boston, SanFran, Houston, etc. Increased NIH support of such core functionality/expertise means:

    1) Shared utilization of the resource(s) by senior and junior investigators alike.
    2) Consistant, expert research without climbing the learning curve or reinventing the wheel each time
    3) Non-duplicative instrumentation costs and reagent pooling. No time and expensive reagent wasting due to inexperience
    4) Cutting-edge modifications of the processes and technologies (always up to date)
    5) Terrific resource for training students and postdocs from a single, reliable source
    6) Career scientific staff of a core are unencumbered by need to publish, ergo data is data and there is no pressure to “create” positive outcomes.
    7) Negative data is archived and reported. Cores KNOW what doesn’t work, so failures (read good ideas that didn’t pan out and thus the failure was never published) are not repeated over and over.
    8) Funding doesn’t benefit just one institution but rather is open to all institutions within a given geographical area.
    9) Research activities accessible at a national core cannot be budgeted to a grantee in the area without additional substantial justification.
    10) Core metrics are easy to measure based upon activity, not based upon publications or “outcomes”. Core expansion is thus need based, core contraction likewise.

    Shared resources being used in truly collaborative scientific pursuit is the only way to repair the research enterprise of the US. The “club mentality” that permeates the current system must be replaced, enhanced core resources are one important facet of that eventuality.

  146. Simply, I totally agree that more than four R01s reduces the PI’s productivity.
    Good idea ! Let more money and opportunities flow in for young(er) scientists.

  147. I support this proposal. It’s incredibly difficult these days for junior investigators to establish careers as independent investigators. This may help mitigate some of this difficulty.

  148. This is largely a most welcome move and critical to the success, sustainability and diversification of our research base, and future research. It will also improve productivity of overstretched and over-committed PIs.

    The wrinkle is with the multi-PI weighting. The proposed scoring system will likely reduce collaborative grants. 50% effort should equate to 50% points.

    The concept that the GSI will limit good ideas, is utter nonsense.

  149. I appreciate that NIH is attempting to deal with real problems due to diminishing resources and increased demands; the current percentage of grants getting funded is creating huge inefficiencies and distress. Some limits are reasonable, but the specifics of the proposed system strongly discourage multidisciplinary/multi-PI research. Also, the large number of points for infrastructure grants ignores that these often support the research activities of many investigators, and probably should not be assigned points. Similarly, training grants should be considered separately and not be assigned points.
    Given that diminished productivity is one of the concerns, the number of grants would seem to be a weak surrogate for productivity, and it would make more sense to impose limits for low productivity, which is being tracked. In principle, the grant review system should be doing this, but additional constraints based on productivity might be useful, acknowledging that the measurement of productivity is imperfect.
    Also, it is clear that we are training more people than can be supported by present research funding, and further cutbacks in training would be appropriate. However, this should be done with a careful analysis by field of training because some fields do not appear to have a surplus of trainees.

  150. I am a mid career scientist that is at a hospital. I have been working collaboratively with other scientists and have some m-PI grants. While my funds are not significant with the current point system I will be limited. As a faculty on soft money the proposed NIH cuts will basically mean that I will have to look for a new job. This new system is not helping but rather destroying careers.

  151. Having been in both large and small laboratories, I find that the smaller ones are far more productive, both in terms of how much science is produced, and how well-prepared the graduate students and post-docs are when they finish. Limiting a PI to 3 total R01 equivalents is a good idea. Having now led a lab myself for a decade, I find it difficult to understand how anyone can effectively manage 4 major projects — there are just not enough hours in the day. And by “effectively manage” I mean bringing value to the NIH by being able to sign off on any manuscript published fully understanding all of the data, and being able to honestly and completely take responsibility for them. Three R01s would be plenty challenging for anyone I have ever met, and many big name labs have only 1 or 2.

  152. By the logic of the proposed system,
    – no Howard Hughes investigator should receive NIH funding since their base award is equivalent to 3-6 RO1s.
    – those who have hard money salary support from their institutions should be penalized with additional points since they can use their grant money for more post-docs and supplies
    – total support rather than just NIH support should be considered in the formula

  153. Unfortunately the Limits on Grant Support to Strengthen the Biomedical Research Workforce does not really explore the unfairness of the process to investigators in todays’ research climate. To make the claim that those are the reasons without evidence that this is what causes the problem, is to ignore science. Please bring investigators and find out from them, their view of what is accounting for having to submit so many applications. As a 30 year investigator, the reason investigators submit several grants is partly due to three main reasons. First the budget cap has not change in the last 10 years or more so to really do clinical trials and have the personnel to conduct them, investigators are only able to charge 5-15% of effort, not 30-40% as it used to be before to have the staff to work on all the components of the trials. Even then these trials are highly subsidized by our institutions. Most of us work in hospitals where we have to generate at least 3/4 of our salary if we want to remain doing clinical research, even as senior investigators. Tenure is a thing of the past. Second, the level of uncertainty in funding is extremely high given a hyper competitive environment, so investigators have to submit several grants without knowledge of which ones will get funded. Third, conducting research today requires a team, it is no longer a solo endeavor. To be able to submit competitive applications and do top quality research requires that you have a team ready to work when announcements come out. It used to be that announcements were given a year or 9 months before they were due. This is no longer the case. Most due dates in many applications have a due date of three months, at most. This means you have to make sure you not only conduct great research but you have the staff to fulfill the tasks you have promised and to publish, and that you can respond to new opportunities in a timely fashion. This requires investing in your young investigators and making sure you can support them when the grants end. The Index really does not take into account that no cost extensions (were you are doing the publications) have small funding and considerable decrease in time. The idea to give 6 points for Co-PIs will eliminate collaborations across sites and multi-site trials. It will be extremely hard to sustain team for transdisciplinary work. I hope you invite the investigators and solve the problem with them rather than for them, causing great harm to the field.
    I have to say that you have squeezed researcher so much in the past ten years that many young people do not think this is a career worth pursuing. This policy will get many senior people to leave the field since the level of compliance, administration, hassles and low budgets is making this work unfeasible and not appreciated. I agree that if researchers are not productive, the should not be granted more funds.

  154. The GSI concept is good but the formula for its computation seems poorly formulated. Providing that a multi-PI designation is valid and that each of the MPIs contributes a fraction of the work, then it is very strange to assign nearly full GSI value to this role. The information provided above explains that the rationale for the GSI is primarily that PIs have only a limited bandwidth to manage research. MPI effort is fractional and therefore the GSI value should also be. For multiple PIs, the sum of their GSI designation should be equal to that for a single PI. Perhaps there will be cases where the roles are not full equal, but the sum of their efforts should not be greater than the role of a single PI – they are the MPI of a SINGLE grant that conducts only the standard amount of work per the funding mechanism.

  155. I think this is well intentioned but needs further thought. At the current time, younger investigators often can not get things funded through study sections, and senior investigators take the PI position to bring the project to the lab, and where funds for the project and junior investigators flow through, This can be a critical source of funding for junior investigators and a critical source of project support. If there is no parallel effort to assure junior investigators can secure good scores through study section, this rule may result in a net diminution of funds to for good projects and actually hurt junior investigators.

  156. Everyone agrees (or should agree) that funding early stage researchers is a good idea, because we are in the process of losing/discouraging the next generation of researchers. The question is how to achieve this goal. As a former study section chair, I do not believe that simply freeing up more money using the GSI method will help, unless there is a formal strategy for diverting this extra money towards ESIs. So like others above, I suggest increasing the “bonus percentage points” for ESIs, from say, 6 to 8, for example. Since we are all competing for the same pool of money, that would automatically shift the funding distribution towards the target population. But at the same time it would also hurt other non-ESIs who may be struggling, so it is not a perfect solution. Others above have suggested a distinct pool of resources for those with no previous RO1 or equivalent NIH funding. But there is already a slight (with an emphasis on slight) existing benefit. These applicants are currently referred to as “new investigators” and they are not necessarily the same as an ESI. They are supposed to be treated more leniently by study sections, but do not receive any official scoring benefit other than at the whim of the reviewers on study section. It would be interesting to see statistics from NIH on the funding benefits of being such a “new investigator”. That being said,I don’t necessarily agree with the separate pool of money for this group, this since many of them in my experience are “established” (using funding from non NIH sources) and they may have no prior NIH funding for good reason – their research is not competitive. In part, this new investigator issue of no prior NIH funding might be resolved using the ESI strategy, since all ESIs are by definition new investigators. Another point is that the scoring system should not inadvertently discourage investigators from participating in team science (by “scoring” PIs of multiple PI grants at almost the same level as individual PI RO1s, because resources are shared) or training (by giving points to those running training grants – the source of our next generation of scientists). The R21 score of 5 also seems too high for a two year, smaller budget grant. Bottom line is that while I agree that some restrictions are needed, the points system is simplistic, is not based on how science really works (some PIs can handle multiple awards and be very productive, some have more difficulty), and may not achieve its goals without a clear strategy defining how reallocation of resources will be achieved by NIH. The simplistic aspect comes from the fact that any given investigator with just a single RO1 but lots of funding from other sources (such as institutional funds, philanthropy, industry and foundations) would not be penalized by the points system. In contrast, another PI with 3 RO1s, but a lesser amount of “additional” funding would be penalized despite having to oversee less work, and fewer people, than the 1 RO1 PI in our example. So if the NIH really believes that more grant money leading to lower productivity is a major argument for applying the GSI index, then other funding sources requiring PI effort should be taken into account. And I don’t think any of us want this to happen or even see how it could be achieved. It’s complicated!

  157. This is a simple solution for a complex problem though it will certainly increase the number of funded investigators. It will address the jobs availability problem. Is that the primary mission of the NIH? NIH’s severest critics call it a jobs program for scientists; this proposal plays directly into that critique. Arguably, the emphasis of NIH funding should be excellence and creativity and intellectual risk taking to advance our scientific basis for medical care. Does this proposal enhance those goals? Are the data used to justify this program accounting for those characteristics or just the mind-numbingly inadequate impact factor? If you selected the 5 papers in each major field in each institute that truly advanced or changed thinking over the last 10 years, were they driven by labs about to be rewarded or those to be penalized?

  158. It seems likely that this policy would hurt rather than help junior investigators. Small grants like R03s and R21s — which should be encouraged as mechanisms to launch burgeoning careers — instead would incur 4 and 5 points, respectively, disproportionately high compared to the 7 points designated for R01s which due to longer duration and larger annual budget will naturally provide many fold more dollars in research support per award. So the point-to-dollar/stability ratio is way off.

    Furthermore, pairing with a more seasoned investigator as a Co-PI on submissions is a common way for more junior investigators to establish a strong team of collaborators to work on their project and to ensure that they can navigate the whole process with benefit from their partnering PI’s experience. But now this is penalized because multi-PI awards don’t split the points equally. Instead each Co-PI is assigned just one less point than if they had been the sole PI.

    The spirit of this policy is honorable but the incentives are completely counter to the stated objectives. Based on the way points would be assigned, the system steers toward submitting large R01s with a single PI. And if those are the types of applications being parsed, that confers a major disadvantage to early- and mid-career scientists competing directly against applications from established investigators.

  159. I have concerns about this new policy and the assumptions leading to its promotion.

    Research showing that productivity does not track linearly with the number of grants does not tell you that capping the number of grants and funding grants that are less meritorious by peer review will lead to increased overall productivity. Firstly, the newly funded grants may not contribute proportionately to research productivity because they will be of lower priority. Secondly, this productivity analysis is not broadly applicable. Some areas of research necessitate a much larger research footprint to do effective research. You may drive researchers in these areas out of academia to the detriment of NIH funded research. It is problematic to compare productivity above the limit and just below, which represent different types of labs, and extrapolate from that comparison that reducing the heavily funded lab to below the limit will only reduce the least productive research. The natural study data do not tell you what happens with intervention. You may end up losing the most productive research in the highly funded lab. Before risking the most impactful research programs, you need to do a focused study to test the assumption that this will help and not hurt. Restricting the 6% of best funded researchers (and overall most productive researchers) based on this extrapolation is dangerous to the overall NIH research enterprise. Furthermore, soft money researchers require more grants to do research than hard money researchers. What effect on research productivity occurs due to non-grant commitments such as teaching requirements or administrative roles? A heavy teaching load seems likely to impact the size of research program that can be managed, yet that is not being considered.

    There are clear issues with the pipeline of new investigators and the age of first RO1. I understand the intent of this initiative and am sympathetic. However I think it is being implemented on the basis of correlation data that compares different researchers at different levels of funding that may be misleading in terms of predicting the effects of reducing the funding of the top funded researchers. There are many correlation analyses of health, for example, that prove not to be causative when tested by intervention. Here you are moving to system wide intervention without adequate data that allows you to actually predict the effect of capping funding.

    Overall, I don’t think you have the data to justify implementing a cap system. Punishing research accomplishments and funding success may be more likely to severely damage high impact research than to augment it.

  160. I think limiting the number of grants is long overdue. In too many cases the NIH is used to supplement operating expenses in an institution and to increase the brand of the investigator. We have PI’s who are deans, treating patients, teaching, traveling the globe doing presentations and lectures, running multiple large scale program projects, and chairing departments and institutes. We have numerous investigators who are PI on 10 or more projects (both NIH and industry/foundation) and are listed as Key Personnel on dozens of additional projects. They could not possibly be meaningful contributors to all the projects they are listed on. And the taxpayer funds this. Some years ago, I sat in on a competitive renewal meeting with a team doing primate research where one team member said several primates had died in the past year. The PI seemed unaware and no one at the table knew the causes of death. While they would be able to pull together the information and I’m sure all required reporting was done, I left the meeting wondering why the death of your research subjects wasn’t important enough to know about off the top of you head. These are not bad people, they’re overextended. Every system needs to change periodically. Otherwise, it gets bogged down and exploited in unintended ways.

  161. Efforts to help young investigators obtain funding are important. But the proposed grant support index is likely to substantially hinder biomedical research progress in the United States.

    The proposal to limit grant support to individual investigators is analogous to restricting the number of patients treated by the most experienced and expert clinicians, to provide more opportunity for less experienced clinicians to see patients. Funding should be awarded to the most innovative, important, and impactful science, especially if it was proposed by the most expert, capable, and experienced investigators.

    NIH has published evidence suggesting that increasing amounts of funding per investigator result in fewer publications per unit of money. However, these analyses may not reflect that more experienced and senior investigators are conducting riskier and more difficult research later in their careers. The comparison between publications per unit funding between early and late career scientists is likely flawed. For example, in clinical research, observational studies are more likely to be performed by young investigators. Observational studies allow opportunities to ‘mine’ the data and produce lots of publications. Later, experienced researchers may be using their observational study findings to conduct randomized trials. Randomized trials are more complex and difficult to carry out. Results are not available until the end of the trial (often years after the start of the study). Because outcomes are all pre-specified, there is not opportunity/ the ability to ‘mine’ the data. If the trial is negative, there are fewer publications. Yet a negative trial is often an extremely important advancement of biomedical science. In summary, more complex, difficult to complete, risky research may yield fewer publications per dollar. This type of research is likely conducted by more senior and experienced investigators. Similarly, the most expert and experienced surgeon may operate on fewer patients and have worse outcomes because he/she is operating on the sickest and most complex patients. This still would not warrant handing off operations to younger less skilled surgeons.

    Furthermore, the metrics on publications per NIH funding unit do not consider influence on clinical practice guidelines. Negative randomized trials can have important implications for clinical practice but yield fewer publications. Yet this type of research is no less important and is more likely to be conducted by the most senior, experienced, and well funded investigators.

    The grant support index will curtail research from the most talented, expert, and experienced scientists in the United States. The proposal is likely to harm junior scientists and post-doctoral fellows who are receiving support from grants of senior scientists. Alternatives methods exist to assist younger investigators. The proposal to limit grant funding to the most successful scientists will most certainly slow the pace of biomedical scientific advances in the United States.

  162. While I support the goal of increasing funding opportunities to junior scientists, there are inherent contradictions to this goal in the new GSI. First, this system discourages Multi-PIs by nearly doubling the points allotted to one grant, although MPI is a way of encouraging and elevating junior scientists. Second, this system discourages PIs from providing infrastructure support (e.g., UM1 grants) for others to do the science. Finally, the assignment of points to a PI of a training grant (e.g., T32, R25), while the PI takes $0 salary from this grant, is a disincentive to support trainees.

  163. Dear Dr. Lauer,
    We would like to share our concerns regarding the recently announced NIH Grant Support Index. This proposed limitation on NIH grant funding will be especially problematic for research divisions like the Channing Division of Network Medicine at Brigham and Women’s Hospital. We would like to raise the following points regarding the proposed Grant Support Index:

    1) Expanding Funding for Early Stage Investigators: We agree that increasing funding to Early Stage Investigators is essential to the future of academic medicine in the USA. However, the proposed limitation on grant funding through the Grant Support Index is just as likely to increase funding to older, less productive investigators who are currently uncompetitive for NIH grant funding as it is to help Early Stage Investigators. We would favor two approaches to help Early Stage Investigators directly: a) Increase the percentile bump that Early Stage Investigators receive for their first R01 beyond the current levels; b) Recognize that obtaining a second R01 by an Early Stage Investigator is often just as challenging as obtaining the first R01, and provide a percentile bump for the second R01 as well as the first. These items would directly benefit Early Stage Investigators without inappropriately reducing support of more senior investigators. Indeed, a key reason for the success of the NIH grant review process is its focus on scientific merit. The GSI proposal goes counter to this basic principle by specifically targeting the most highly productive and meritorious scientists, whose proposals have withstood the scrutiny of peer review.

    2) Unintended Consequences of the Grant Support Index: As currently proposed, we are concerned about several unintended consequences of the Grant Support Index:
    a) Reduction in Use of the Multiple PI Mechanism: We have found the MPI mechanism to be very helpful to acknowledge the important role of junior investigators, and provide them guidance and support for their second and subsequent R01 grants (we avoid use of the MPI mechanism for first time R01 recipients to protect their ESI status). The MPI approach has also increased collaboration within and between institutions; we are frequently contacted by investigators from other institutions about developing collaborative MPI projects. If limited GSI points are available, this could yield investigators who are more focused on individual, internal research projects and have less room to play a key role in external collaborations.
    b) Premature Termination of Projects Requiring a No-Cost Extension: It is not uncommon for NIH-funded research projects to require a No-Cost Extension to complete the proposed research. However, if GSI points are expended during a No-Cost Extension, investigators will have strong incentive to terminate those projects before the work is completed in order to preserve their GSI points. This could lead to suboptimal use of NIH funds, and to rushed, less rigorous research.
    c) Penalties for Institutional Grants: Although no salary support for the PI is included, the current GSI proposal includes points for leading a T32 grant. This will discourage leadership of this key training mechanism, and directly penalize young investigators. In addition, the large proposed point allocation for UM1 grants (11) would markedly reduce the incentive for investigators to lead those critical infrastructure projects.
    d) Penalties for Leading Subprojects under Multi-Component Awards: If each Project Leader or Core Leader in a P01 grant incurs 6 points in the GSI, it will markedly reduce incentive to participate in those important collaborative projects.
    e) Less Transitional Support for Junior Investigators: Senior investigators with multiple R01 grants use those funds to support junior investigators who are transitioning from K awards to their own R01 funding. Numerous faculty at the Channing Division of Network Medicine have been able to launch successful independent careers through transitional roles as co-investigators on grants led by senior faculty. Without this support to senior investigators, we contend that the number of successful junior investigators in academic medicine will actually decrease with the GSI approach.

    In order to achieve the desired goals of empowering Early Stage Investigators without producing the negative consequences noted above, we would suggest implementing changes in grant scoring directed at helping Early Stage Investigators obtain their first and second R01 grants. We do not support the use of the GSI. However, if the GSI is to be implemented, we would strongly favor removing points for T32 grants, reducing points during a No-Cost Extension, reducing the points for Multiple PI grants, reducing the points for institutional grants like UM1, and eliminating points for individual Project Leaders or Core Leaders in P01 grants.

    Sincerely,

    Edwin K. Silverman, M.D., Ph.D.
    Scott T. Weiss, M.D., M.S.
    Meir J. Stampfer, M.D., Dr.P.H.
    Fran Grodstein, Sc.D.

  164. I fully support the gradient payline strategy. It should be a better fair play if we categorize PIs into more different buckets based on their current funding situations and stage of the research career (e.g., how many R01 obtained before) and let them compete within the same bucket. It is not fair to ask a junior PI with one student in the lab having no money to compete with a mega lab on writing a grant. Clearly NIH not only needs to sponsor the best science, but also needs to foster the next generation young scientists.
    For the ESI with no prior R01, the payline should be at least 20-30 for this pool of people to allow more people to be in the play.
    For established with no current R01, the payline should be 20 for this pool of people
    For established with one R01, the payline can be 10-20.
    For established with more than 3 R01, the payline should be within 5%

    In that way, we dont’ kill the best science while enable junior people or people desperately need one R01 to survive.

  165. I am in support of the GSI. I think there is too much of a club and as a junior investigator, it is tiring to watch people have multiple R01s, P01s, etc. and feel like one can never get funded as a new person with a different idea. The assumption from my institutions has been that the senior scientists will share resources and mentor people as well as “grow the field.” That doesn’t happen in many, many cases. Often the well-funded researchers refuse to meet with you, ignore written correspondences repeatedly, and share absolutely nothing with junior people, acting as though they haven’t a dime to invest in someone young or new. You are on your own as a junior person in a lot of institutions trying to get funded and surrounded by heavily funded researchers who are judging you by how many grants you hold. Enough is enough.

  166. This topic that has bothered several small investigators for last 2 decades. The proposed changes might stop the club culture that has proliferated the research realm. From my experience of last 17 years, I have come to one conclusion; big labs are mainly interested in garnering funds from every possible sources including small foundations to generate many patents, their medals. There is no interest in mentoring, sharing resources with less fortunate investigators or advancing a common goal.

    Not only should the funding be restricted but also these “too big to fail” labs should be encouraged to collaborate with other investigators in different labs in country or overseas who are also exploring similar ideas. In the current state, it doesn’t even matter how novel idea is being proposed or investigators talent and experience, if you are not part of a big lab there’s no way on can succeed in securing funds (sharing a collective experience of our group). The system is completely rigged in favor of self selected few. I have been on few study sections and I have observed that basic scientist have an agenda and often kill very interesting clinical proposal in favor of those with more emphasis on the basic science. Unfortunately, these efforts are slowing killing the reason NIH was created. The NIH should completely revamp its system including the reviewer pool to disrupt the country club culture. Also, experienced clinical investigators should be part of the NIH reviewer pool. The proposed changes will lead to better ideas and more useful research outcomes.

  167. As expected, this proposed change has polarized the research community. I’ve always supported the idea that we should fund the best science, period. Reading this blog, one would get the impression that peer review is broken, well-funded scientists are hoarders. intramural NIH is bloated, institutional indirects are bleeding the system dry, the playing field is not only uneven but riddled with holes and land mines, and on and on. After being in the system for 30 years, my feeling is that what’s happening is a fairly predictable and recurring event for those who have been around long enough to remember.

    Similar concerns and feelings were voiced during the Reagan administration when NIH funding was in decline, pay lines dropped into single digits, and researchers became demoralized. When there was enough money in the system and NIH was funding closer to 20th percentile (1 in 5 applications), people were fairly satisfied that good science was getting supported. When funding drops to 10% (1 in 10 applications) people become frustrated, and when the funding drops into single digits (below 10%) the discontent transforms into anger. I’m fairly certain that if more money were in the system, rather than 20% cuts proposed by the current administration, that none of this realignment would be discussed or necessary. More money would almost certainly fix the problem. But we are here and this is the reality unless priorities in the White House and Congress are radically changed. So now what?

    A few of my opinions:

    1) Existing policies regulating percent effort, targeted programs designed to favor early career researchers such as handicaps in priority scoring, specific programs to foster independence as intended by K awards, training programs, and assessments of negotiated institutional indirect are all in place and could be modified or expanded to achieve much of the same goals as this proposed realignment.

    2) Assigning points to PIs for training programs such as T32 or U54 will drive any research-active faculty out of managing such programs, as these types of funding bring nothing in terms of laboratory support to the PI. If the policies are enacted as currently proposed, then I would be forced to withdraw participation in training grants altogether to focus solely upon using points to maintain my laboratory. If anything, participation in training grants should be a credit for additional points above 21, since it is working to help early career individuals as is the purported intent of these changes.

    3) Having multiple policies or sliding scale formulas for soft versus hard money positions will rapidly devolve into a quagmire. Establishing minimum standards of salary support for researchers from the institutions they help to support with their funding and the associated indirect seems reasonable.

    4) Assessments of how many trainees a lab supports and the record of accomplishment for a lab in producing the next generation of researchers could be an independent metric to help determine if additional support for the lab in terms of available points or a higher funding cap is warranted. These productivity metrics would have to be re-upped every 3-5 years as individuals moved in, through, and out of the lab to demonstrate continued fostering of new researchers.

    5) Sadly, there will be some researchers whom, for whatever reasons, are not capable of creating and maintaining a competitive research program. We all knew the system was competitive when we signed on for the career, so to blame the system or those who are successful for one’s inability to maintain a research program could be perceived as bitterness or sour grapes. Many of my colleagues who are terrific scientists are struggling and it is a shame that more support isn’t available and that the public does not pressure Congress for more NIH support considering it is such a small percentage of the Federal budget. I’d like the community to not lose sight of that reality and try to recognize that our real goal should be to increase funding for biomedical research, not turn on one another in frustration, jealousy, or envy and cannibalize many of the long-standing carefully crafted programs and policies that have evolved over time. Many laboratories (both excellent and marginal) across the country will be hurt by GSI implementation just as many laboratories will benefit. Who deserves which fate is a tough judgement call. The GSI will not be a panacea for the issues we are struggling with – it will trade one set of problems for another set of problems.

    In the end, revolutionary change often comes at a heavy price. The GSI point system and imposed caps on funding feels draconian and at the far end of a pendulum swing. Sometimes the pendulum is coming towards you, eventually the momentum takes it away from you, and if you wait long enough it swings back in your direction if you’re patient enough to stay put. For those who have been around long enough to know better and live through the pendulum swings, carefully executed moderate adjustments and gentle repositioning is a more effective approach to staying on course rather than tearing up the status quo and risking driving into a ditch. The more we fight amongst ourselves, the less inspired the public will be to support our research enterprise. Everyone on this blog presumably wants science to grow, for good science to get rewarded, for great mentors to keep training the next generation, and for the public to appreciate and respect the contributions we are making toward improving human health and quality of life. So let’s try to keep this introspective assessment as productive, respectful, evolved, and mutually supportive as we can for our own good.

    1. I am in complete agreement with this post. The proposed GSI system penalizes experienced researchers who are the mentors of the young investigators that we are trying to prepare for independence. If a researcher is successful at being creative and developing ideas that pass peer review, it seems they can no longer be funded because of their success in doing good science. The number of grants one has received is directly related to the age of researcher. Should NIH be supporting ageism? I don’t think so. Discrimination in any form (ethnicity, gender or age) should be adamantly discouraged. Moreover, as “old enough to know better” notes the government will get the most for their tax dollar when we simply fund the best science no matter who is doing it!

      I am definitely in favor of supporting young investigators through administration of handicaps where new investigators are recognized at the time of grant review or by the Institutes providing funding. We need more resources put into K awards and training programs and less in Centers and Collaborative awards (U01s) that are very expensive. The best accounting of our research dollars comes from programs that are small enough that each investigators must be accountable for what they produce. We all know of large scale programs where some researchers are highly productivity while other researchers are along for the ride. Why can’t we investigate these and other ways to save research dollars for young investigators rather than assuming everyone with past histories of grant-getting success should quit doing the very thing they are good at doing.

  168. As a junior investigator myself, I applaud for and benefit from NIH’s initiatives for increasing the odds of successful grant applications for new investigators, such as better paylines for K awards and better paylines for first R01 of junior investigators. NIH shall definitely continue to work on supporting junior investigators, although applying a GSI will not help them, at least not all of them. The NIH fails to consider the complexity of salary systems at US research institutions and universities, probably because NIH leaders and scientists never worry about getting paid by the federal government. Many faculty/research positions at US universities, especially some top ones, are soft money-based. This means that investigators are expected to pay 100% of their own salaries. NIH grants are no doubt the primary, if not the only, source of funding. Three R01 grants are not going to cover 100% salaries. That’s the reason why junior investigators are often work very closely with senior investigators on senior people’s projects to get the rest of FTE covered AND to get some experience about how to run a project and how to supervise a postdoc, and etc. When senior people are getting three R01 themselves, how could they even possibly support junior investigators? Do the math. NIH salary cap is roughly $180000, so for each R01 senior investigators will have to budget for $60000 each year. That’s a postdoc’s full salary. After deducting salaries for research staff and lab costs, there’s little room in the budget to cover salaries of postdocs/junior faculty. And getting three R01s is very hard for junior investigators. Then how could junior investigators survive and let alone support their own postdocs? I foresee that when this policy is really in place, senior investigators will be forced to still write the grants and run the projects as the true PI at backstage, and just use junior people’s names in the applications. Is this what NIH really wants? The other consequence is that we will have to seek other sources of funding, including the industry. How does a wide-spread COI with industry help with research in the US? Many good investigators may have to change their career to the industry paths. I think we all need to remember how important a merit-based grant system is to the scientific research in this country. The tax dollars shall be invested on the best research and best people who can make that research happen.

    And if NIH really needs something to do, how about improving the peer-review system? I was lately often shocked by the irresponsible comments raised by reviewers who did not carefully review my grants, for which I spent months in putting everything together. Will NIH implement a policy to disqualify bad reviewers from study sections? Will NIH consider policies to encourage investigators with existing funding experience to help judge other investigators’ grant proposals? To me, these are way more important than the GSI, which is by no means a very thoughtful way to improve funding situations in the US.

  169. Although many others may have already weighed in with similar points, here are my concerns:

    1. The point system will disproportionately affect “soft money” institutions and labs without the discretionary resources to provide salary support to its investigators from non-NIH sources. The impact on training, recruitment, and retention would be devastating.
    2. The system would likely have a severe impact on collaborative grants, which are essential for any sort of big data approaches or intervention studies. It would also reduce collaborations on R01 and other similar grants.
    3. T32s should be exempted from this point system unless the PI is taking salary. I don’t see how reducing the number of training awards is helpful to the mission of the NIH.

  170. I am an established investigator strongly in favor of more funding for junior and mid-career investigators. However, the proposed system is too “one size fits all” without taking into account several important factors. The new system should take into account the following issues.

    1. Investigators devoting 90-100% effort to research at top institutions that provide little or no salary support will be unfairly disadvantaged; many of these investigators are productive scientists running the best labs in their field, and publishing high-impact work that drives the field forward. These PIs often devote much time and resources to training and mentoring junior investigators, and will be less likely to do so when funding is capped at 3 R01s. Many of these investigators have shown they can successfully manage more than 3 projects, are often involved in multi-PI projects, and often have experienced staff scientists in their labs, with higher than average costs. Capping funding for investigators devoting full-time effort to research at the same level as investigators who receive substantial % salary from their institutions for teaching or clinical work will be detrimental to many outstanding labs, and will have downstream negative impact on junior investigators who benefit from training in these labs.

    2. Funding gaps are a big problem at soft-money institutions where there are no funds to cover them. We need mechanisms to prevent “falling off the cliff” when grants end around the same time. PIs don’t know when their next grant will be funded, and may apply for several grants around the same time, not knowing whether they will get funded the first time or after several submissions. The system should be flexible enough to accommodate temporary “bumps” in funding, which are critical to prevent future “drops” that lead to closing down labs or losing experienced personnel. Mechanisms to “smooth out ” funding levels over time and prevent funding gaps could include: 1) NCEs without GSI points; 2) allowing 6 or 12 months at higher GSI; 3) higher GSI limits (e.g. 4 R01 equivalent to accommodate additional smaller or multi-PI grants); 4) flexible start dates so a new grant can begin when an existing grant ends.

    3. No-cost extensions with little funds remaining are a cost-effective way to finish challenging projects and avoid funding gaps; there should be little or no points for NCEs.

    4. Other sources of NIH funding that are likely to provide more value to science than capping productive labs at 3 R01s: 1) shave 10-15% off IDC rates, which are disproportionately high at some institutions. IDC capped at 50-60% would free up significant funds for research. 2) cut budgets on costly centers, consortia, and large special projects – many of these costly endeavors do not yield as much productivity per $ as R01s; 3) cap NIH intramural labs at similar levels as extramural labs; 4) require institutions to cover at least 10-20% salary for research faculty.

    5. T32s and other types of service such as leading a center is critical for science, and should not count toward GSI; otherwise, excellent scientists will not be willing to perform these roles.

    6. R21s and other small grants should get fewer points than proposed, given low budgets and short duration.

    7. 3 R01 limit is too low for producing the best science; a 4 R01 limit is more realistic given that many of the best labs are functioning at or above this level.

  171. There is definitely merit to limiting the amount of funds (or grants) that individual investigator can get.
    1. I have major concerns about “penalizing” PIs who elect to take time to serve as directors of training grants. T32 grants should probably have a score of in the negative integers perhaps -2 or -5 in a GSI scoring system.
    2. More thought needs to be given to multi PI grants and to help people are going to game the system by just submitting less grants of larger magnitude.
    3. NIH might also consider changing the payline for a PI’s fourth, or fifth, R01 grant, perhaps requiring a much lower (better) score, eg., 2nd or 3rd percentile, in order for that grant to be funded.

  172. My first impression is that this approach will not solve the problem in the ling run, as highly motivated successful PIs will find loopholes to get research money: applying for 400,000 R01s, getting junior research assistant professors as PIs, etc, while decreasing intercollaborative grants, for which the points might be too much to be worthy. In the short run however, many big labs will need to downsize, with the disruption it will cause to the junior investigators in these labs.
    I always thought and expressed this to several NIH officers that one of the best ways to solve the young investigator problem is to make mandatory in multiple PIs grants, in P01s, in U19s, in any grant with more than one PI, that at least one of the PIs is a young investigator. Since there are a lot of these initiatives, this will provide with a source of young PI funding, increase collaborations, and incentivize institutions to hire young investigators before they get their first grant. I bet one gets as many new investigators grants this way as with the proposed cap, but without causing a disruption.

  173. 1. Will disproportionately affect “soft money” institutions and labs without the discretionary resources to provide salary support to its investigators from non-NIH sources. The impact on training, recruitment, and retention would be devastating.

    2. This will eliminate jobs as less money will be available to support the personnel

    3. The math tells you that imposing the point system will in no way and shape cover the salary cap of the investigators.

    4. Less money available to support Science and push Innovations

    5. The proposed system would likely have a severe impact on collaborative grants, which are essential for any sort of big data approaches or intervention studies. It would also reduce collaborations on R01 and other similar grants. It forces you to think small.

    6. Wealth redistribution is artificial

    7. If this is going to be implemented which I hope should not you need to eliminate points for individual Project Leaders or Core Leaders in P01 grants

  174. I am the PI of a T32 training grant in Systems Biology. I do this job not for any personal benefit, but because I believe in the science and the training program benefits PhD trainees and younger program colleagues who recruit the students to their groups. This activity takes time, but it seem unfair to count it against my research award cap. Our program needs to replace the training grant PI every few years to ensure we maintain energy and bring in new ideas. Also, frankly, to share the administrative burden. If training grants count towards the research cap, why should any of my colleagues agree to take this job on? In particular, it will become more difficult to persuade our most research-active faculty to take their turn at the T32 PI job, though they may be the best qualified. I worry the losers in this will be the PhD students.

  175. The grant support index is long overdue. As a young investigator, I continue to see senior investigators receive R01s as PIs on topics that they may not even have real or full expertise on but seem to be a safer bet to receive funding, because they have multiple grants already or know the right people to get invited to planning meetings. Then they depend on junior investigators to do the work or know the topic fully, including at times write parts of the grant. When the grant is awarded, the young investigator gets cut off of the grant funding entirely, and one cannot do anything about it. Then the R01 gets supplemental funding, and still the junior people get nothing. I have been listed on 3 funded Rs as a junior person but cut off each time the grant was funded.

    1. Precisely – amazing the level of hypocrisy of the “super stars” around here. It is pretty obvious that super stars being so busy schmoozing in conferences have no time for creative stardom thinking…more likely 10% of their 35 postdocs provide all the shinning and those should be getting RO1s instead.

  176. As an ESI working towards my first R01, I am very concerned about the impact of this policy on collaborative science, specifically because it is not clear how multi-PI grants will be weighted. As alluded to in many comments above, I and other junior scientists often choose to work with senior mentors who are also our co-PIs, both because the experienced personnel help reassure the study section and because they bring vital wisdom to the research. Those of us who do tools/methods development also naturally end up as co-investigators on others’ grants. If multi-PI grants assign a substantial number of GSI points to investigators other than the contact PI, it would make it nearly impossible to simultaneously run one’s own lab AND get true credit for enabling others’ success.

    Many possibilities would be better, including:
    – Dividing the GSI points for an MPI grant among the named investigators, with an allocation that emphasizes points for the contact PI
    – Allocating GSI points in proportion to the personnel effort (investigator and/or trainees) going to a specific lab on an MPI award
    – Allowing investigators to specify, as part of the Leadership & Management Plan, exactly how GSI should be allocated among the key personnel

    I have a longer blog post which could be found on the Academics for the Future of Science website that lays out the examples of how and why this would be detrimental, both to me and to my peers.

  177. In its current state, the GSI really damages collaborative science.

    My suggestions are: 1- that the initial rollout of this plan focus primarily on R01 and R21 grants that are not ‘team science’.

    2-Please strongly consider that Multi-PI R01, U01, P01 and similar types of collaborative grants be either exempt from the GSI entirely, or be exempt initially and then phased in over time at a much later date. 3-If such collaborative grants are ultimately phased in, please consider allotting them a lower GSI score, particularly for subprojects.

    As it stands now, including these types of collaborative grants into a scientist’s GSI ultimately limits everyone’s ability (both >21 and <21 groups) to collaborate, as no scientist can be 'hit' with 5 or 6 points for a part of a project that brings in research income less than an R01. We can't do it from a laboratory fiscal perspective, despite our being interested in the collaboration. Fiscally, as science is so expensive to perform, we would all have to focus on the sole-PI R01, which then prevents collaboration.

    I think we all understand the concept that the GSI Rule is trying to achieve, but in its current state, it really damages, and ultimately prohibits, the whole concept of collaboration.

    1. @Catherine You do not need a formal stamp of collaboration through a grant in order to collaborate. I have collaborated on several projects and it was informal, with the understanding that collaborating authors will be on the paper. Some people may have 10 collaborations, which is ridiculous. Realistically, how much can you contribute to each project when you have 10 collaborative grants? The problem here is that investigators are really after the grant money, not after the actual collaboration, and take advantage of the NIH idea of promoting collaboration by submitting grants which are technically collaborative, but may or may not be collaborative in reality. If you want to collaborate, you can simply pool the resources that are available to each individual investigator.

      1. Gabriel, what you wrote is nice in theory but not always practical. What if there’s an investigator who needs guaranteed and foreseeable funding to commit to a particular collaboration? At certain levels, you do seem to need this type of guarantee.

        Also, it’s possible for some PIs, with widely applicable technologies/techniques, to be able to collaborate with 5-10 groups at the same time and not lose their core focus.

        Not all research is conducted the same way and with the same constraints. It’s important to be mindful of this heterogeneity when thinking about the current proposal and others.

  178. I am deeply troubled by the proposed limitations on numbers of RO1 grants that an individual investigator may accept. I do understand the need to save money, but imposing what is in effect a quota system is both unfair and unwise. What NIH needs is to advance progress in biomedical research as far and as rapidly as possible. My view is that success in this endeavor is a product of wonderfully smart and beautifully trained investigators who conceive great ideas that lead to powerful solutions. In today’s environment, independent investigators must also collaborate intensively, because they must be ready to capture large, extraordinary research opportunities in the interest of making major research advances. Success in this endeavor requires both individual successes and successes that come from intensive, collaborative interactions among interested and highly competent investigators.

    Thus, what I believe is most needed to fulfill these requirements is healthy maintenance of the work of the largest collection of exceedingly successful, NIH-funded investigators. Cutting back on an opportunity to earn NIH support beyond a limited, government- mandated level will create a climate of frustration not to speak of suboptimal accomplishment-exactly the opposite of what is needed.

    The US needs more top investigators working successfully and together to unravel the great medical and biological mysteries at hand and thereby to reward society with all of the health benefits such discoveries can generate. A points/quota system will do the opposite, not the least of which will be to dampen interest among our young students of medicine in pursuing careers as physician-scientists. This career track is already winnowing and in jeopardy of dying. Let us not hasten its decline. Such an outcome will disadvantage medical research even more severely in the years ahead.

    Therefore, I believe that we must continue to reward the best scientists, who write the most appealing and highly rated grant applications, and whose records of accomplishment befit the level of their funding. We must not impose a limit on the number of highly rated RO1 grants each can accept. The country and biomedical research, in general, need every ounce and not less of their effort.

    If the points system were imposed, investigators everywhere may seek to apply various forms of creative gamesmanship to apply for RO1 grants. They will, in part, be aimed at overcoming the limitations imposed by the quota, using unfair methodologies. This is likely to add a source of logistical corruption to RO1 competitions, which is last thing the NIH needs.

    Finally, if one wants to maximize the distribution of NIH funds to the most deserving recipients, my sense is that the answer lies in reforming the peer review system. It must be vastly improved so that only highly accomplished, constructively critical, fair-minded, and truly insightful reviewers be invited to serve on NIH study sections. Study section chairs must be the best of the best of this group of reviewers and be selected by a formal process led by top scientists in each relevant field.

    And more than 2 RO1s and the formal backing of the top people in the field covered by such a study section’s grant proposals should mandate that an investigator must serve on a study section for the length of at least one of the two RO1s.

    1. @Livingstone Grant limits are necessary in order to prevent abuses by a limited group of investigators which hoard all the grants in the name of “high quality research”. Have you ever thought that maybe a lot of scientists have great ideas but are prevented from bringing them to reality because a group of “great labs” form a protective layer around the grant money, making it very difficult for everyone else to get through? Peer review is often a joke, as you pat my back and I pat yours. This reminds me of a conference I went to years ago in Eastern Europe (during communism), when at the final banquet, participants from some communist countries positioned themselves in a strategic circle around the banquet table, so that it was very difficult for everyone else to have access to food. This image suggests very well how NIH grant funding works.

      1. Gabriel R is saying as it should be said – honestly, and bravely. There is too much corruption, born out of greed and a desire to cling on to power, by the fat cats who hold a huge number of grants. PEOPLE, INCLUDING THE GENERAL PUBLIC, ARE QUICKLY LOSING FAITH IN THIS PLUTOCRATIC SYSTEM. To NIH Officials: Drain the swamp. If you don’t act, don’t complain when your own constituents don’t support you when the budget falls under the ax.

  179. I am deeply concerned about the damage that the GSI would have on research in the US. I see several major issues, which have been mentioned already by others. For emphasis, I will focus on one matter here:

    The new rule would kill collaborations.

    Investigators cannot continue to collaborate efficiently if given nearly the same amount of GSI points as co-PIs, but only ~half the $ amount.

    The GSI seems to have several fatal flaws; its destructive impact on collaborative research is one of them.

    1. @Mikael You are right about the generation of too many post-docs. But that is also the doing of large labs who need extra hands to work on the exceedingly large number of projects generated by too many grants in one hand. Such labs not only promote the acceptance of many graduate students, but also often import post-docs from overseas in order to fill the positions, creating an additional excess of personnel. Furthermore, PIs who cannot get a grant stimulate the acceptance of PhD students in their departments, since grad students are paid by the department and they get free labor. In a way, the number of students interested in a PhD is self-regulating by the ability or inability to find jobs. However, this is a very slow mechanism. For people who started a PhD 15-20 years ago and did not find a job yet, it is too late to change careers. But people who are now considering a career are clearly steering away from research because of lack of jobs/funding, which will lead to a self-perpetuating cycle of bringing in more foreign scientists, which will in turn be speculated by politicians and large companies. 2 R01s will limit the number of personnel that each PI can hire, and automatically lead to a decrease in the number of PhD students accepted by each department, ensuring that only the best and the most passionate about science get to do science. Also, more PIs would be able to hire post-docs with some experience and their interest in accepting a large number of grad students would diminish.

  180. Imposing an arbitrary limit on the amount of NIH funding will not mitigate the challenges facing biomedical research community today. It will certainly spread money, but will negatively impact discovery. Would an NBA coach tell the player to limit 21 baskets per game? People are already thinking about appointing their postdocs as research assistant professors so that they can apply for NIH funding. Is this the right path?

    I think that we are producing too many PhDs for too few real jobs. That is one of the reasons that graduates have to do many years of postdoc research. Too many scientists compete for the ever shrinking pool of fund. Federal funding has remained flat in the past decade, while costs have increased significantly simply due to inflation. The oversupply of PhDs happened to other fields (for example, chemistry and physics) decades ago. Now those fields don’t face the same crisis as the biomedical field.

  181. As a young investigator in her fourth year as an assistant professor, I have just now reached the time to evaluate which of my projects are worth continuing and which ones are not; I have established expensive but very useful animal models of disease to generate data for both current and future NIH proposals. I have hired and trained staff, students and post-docs. Because I worked extremely hard these past few years, and took many grant writing classes and asked to be mentored by my senior colleagues, I got out 60 grant proposals (NIH and private funding or foundation grants, pilot grants) since 2013, and I now have 2 RO1s, 1 R56 and a private cancer grant. I currently have a grant pending as an MPI. I am also required to bring in at least 60% of my salary, which is low compared to most of my colleagues, who have to bring in 100% in their academic departments (I am in a clinical department).

    In my opinion, should the point system go into effect, the following consequences will significantly damage my career:

    1) Should I get the pending grant, I will not be able to re-submit my R56 as an R01 because I am maxed out (unless I refuse to accept the pending grant, which will obviously cause damage to that collaboration). I already have a post-doc hired, and for the pending grant, anyone can easily learn the already established skills we use every day to perform these tasks. Thus, I disagree that it would be unproductive for me to have both grants. I spend significant time setting up both systems, and mentoring my staff, unlike some PIs who have Biotech companies on the side.
    2) I have spent my startup on high-risk research, which is now ready to publish, with interest from high impact journals. I believe that without this publications, I would not be able to get immunology grants on these rather novel findings that do not reflect textbook teachings and are thus not good candidates for NIH funding until backed up by a peer reviewed publication. With my points maxed out, I am now considering to not publish these result as I cannot apply for funding on this topic any more and as such would not be able to do any follow up work. Publishing the result in this case would only allow others with funding or points left to take over the project from my lab. If I did want to fund this project, I would have to dump a current RO1, and then have to risk not getting my new grant on the new topic while not renewing my parental RO1 on a topic the NIH and my lab has already heavily invested in. I also have spent significant time establishing the needed expertise in my lab. It would be wasteful and counterproductive to discontinue HIV research. In addition, I already successfully mentored on both projects so clearly, I can be efficient and productive on both projects.
    3) With my startup money gone, I will have to fire staff and be understaffed for my current projects.
    4) I would reduce the amount of in vivo science in my lab because I cannot afford it. Thus, I would spend money on mediocre science and papers rather than work on groundbreaking discoveries. I suspect that I would find myself very frustrated by that.
    5) I have already told two junior investigators that I will not be submitting the grants with them that we had already started writing and planned on submitting this fall, because I do not want to spare any points. I will not do MPI grants where I have to share budgets but get almost the same number of points as a single PI grant. I will not participate in PO or U grants.
    6) I have approached industry to fund some of my science (risky and restrictive, in my opinion). I suspect that private funding will be very hard to come by now that all senior established PIs will apply for those grants to make up for lack of NIH funding.

    I would rather propose to have a minimum and a maximum cap of effort someone needs to put on a grant, as there are many senior PIs who have 0.3% effort on 20 different proposals etc. This, clearly, is not productive. If one had to put a minimum of 5%, then real collaborations with effort would remain, and the CV stuffing would stop. One could review whether PIs with Biotech startups should have a reduced grant load, as their efforts may go elsewhere. One could put a minimum number of papers need to be allowed to apply for a competitive renewal. There are other ways to do this.

  182. The point system that the NIH plans to implement will be detrimental to PIs involved in community projects as in most cases they run active research labs. The NIH should look into this issue seriously as it will discourage PIs to be community resource builders.

  183. I am worried about the allocations of these points….why would being a T32 PI, for which you can’t allocate any salary for yourself, make you unable to have 3 R01s?? This is the case with the current system of points. This makes no sense, as being a T32 PI is not doing research-you are just helping trainees find labs of other PIs on the training grant to do research. This does not take anywhere near as much time as directing research on an R01 grant.

  184. The small percentage of labs with larger amounts of funding will be irreversibly hurt by this poorly conceived scheme. These labs A) disproportionately train future investigators, even when adjusted for size B) disproportionately fuel high risk innovation C) provide a resource for protocols, reagents, and technologies that is only possible because of sufficient size and stability.
    The US has led in biomedical research by funding proposals based on merit, not on central command-and-control planning to spread resources around. This will devastate US science. Every proposal funded has to run a gauntlet of peer review, that implicitly incorporates the productivity trajectory of the individual investigator. If proposals can be funded even after that sort of scrutiny, should they be denied by a catch all mechanism like this?
    Moreover, do you really want people to abandon T32 PI leadership when they have to give up an R01 to do so? That is what will happen and it will severely hurt the next generation.
    While young investigators need funding, this is not the way to secure it. Elimination of excessive RFAs that divert funds from investigator-initiated projects is a far wiser way to go, albeit one that might not be popular with Washington bureaucrats disconnected from extramural science who always seem to want to put their “mark” on NIH.

  185. I am a junior faculty with a modest lab, but the proposed GSI score would be debilitating to my lab because of the way it penalizes collaborations. I would have thought that collaborative labs should be encouraged, perhaps getting more funding rather than being eliminated.

    This point has been made before – I want to add my personal experience. I am generally in favor of the GSI cap, except for the handling of T32 training grants (as Tim Mitchison points out), and of collaborations.

    I am an Assistant Professor in the field of Systems Biology, with major efforts in technology development (single cell sequencing), and in applying computational approaches in the biological sciences (stochastic processes and bioinformatics). I am in the process of submitting collaborative R01 proposals with collaborators having entirely complementary skills sets (hematopoiesis, immunology). I am asking for under 50% of grant support to cover our role. Despite my low “cut”, we are central to the inception, design and implementation of the grant proposal, and should be awarded co-PI status.

    These collaborations bring together experts in totally different areas. I cannot see why the NIH would want to penalize us with 6 GSI points for less than one R01-equivalent of funding in total across all grants. To be clear: if all grants are accepted, I would be left without enough points for even a single R01. Doing collaborative work and fostering team spirit is risky enough even with full funding – but this would be totally unsustainable.

    If this policy comes into play, I cannot see any alternative but to cut off or heavily restructure all collaborative efforts. The way the GSI penalizes collaborations seems to be the single most damaging aspect of the entire GSI initiative, but also the one most easily fixed. Some suggestions:
    1. For multi-PI grants, one can simply distribute the 7 GSI points among PIs on a pro-rata basis.
    2. In addition, to actively encourage collaborations, have the *entire* R01 equal 6 GSI points if it includes multiple PIs (say with minimum 20% per PI to avoid fakes). This means that collaborative labs could in fact get more funding, not less.
    3. In addition, since the cost of different components of the grant can be unequal, one could give the PIs the option to decide on how to they will distribute the points among themselves.

  186. I am very concerned that PIs of training grants will have points given so that they can’t potentially have 3 R01 grants. There is no money on training grants to support the PI’s salary. Most PIs run training grants to help colleagues and PhD students develop research programs. The points for being a training grant director should be removed. There is some administrative burden associated with being a PI of a training grant but it is not much compared to running a research grant!!

  187. A major change in NIH funding policy such as “the rule of 21 GSI cap” is likely to have both intended and unintended consequences. The first intended goal is to get more “bang for the buck.” It is not at all clear that limiting support for highly funded investigators will increase quality or quantity of output in biomedical science. The published analyses presented as justification are simply not convincing and the use of log scale graphs is belied by actual numbers of publications linked to 4th and 5th R01s vs. first R01s. If implemented, this will be a high risk experiment. The second intended goal is to change the demographics of the research community to increase the percentage of funding going to early stage investigators and mid-career investigators in order to stabilize the work force. No analysis was provided to link this issue to the proposed solution.
    Among the most likely unintended consequences of the proposed rules are diminished collaboration due to the high penalty set for PIs involved in MPI applications and the penalty imposed for the service of running a T32 training grant. Both activities will clearly be discouraged. Can we rethink this? Are there better ways to support early stage and mid-career scientists. For example, one could aid younger investigators with one R01 by increasing the size of the R01 budget and the length of the period of grant support. In addition, one could encourage senior investigators to consolidate their support by increasing the R35 program which asks established investigators (3 or more R01’s) to accept 1 very substantial grant ($750K-1M) with a 7-10 year time frames. If NIH is already committed to imposing such the GSI system, at the very least, the points assigned to MPI grants should be decreased and PI’s of T32’s may even deserve a credit. It’s a healthy sign that NIH has asked for feedback, but were there any R01-funded investigators in the room when this policy was formulated?

  188. I have never had more than 3 R01 equivalents but still believe that the cap proposed by the NIH is a misguided policy for the following reasons:
    1. It will adversely impact the chances that major discoveries will be made by punishing the most successful (very un-American, isn’t it?)
    2. It curtails funding to labs that deliver the lion share of training the next generation of scientists.
    3. It limits collaboration.
    4. It is based on misinterpretation of the data. See commentary by Shane Crotty at La Jolla
    5. There are much better ways of finding money to fund new investigators; for example, by diverting money from the burgeoning share of NIH $$ going to non-R01 initiatives, some of which have had dubious success.
    6. The word on the street that it is a ‘politicized’ rather than a rational decision. The scientific community should not acquiesce to decisions made in this manner.
    Thank you.

  189. I think the intent is good but the plan flawed. Rather than having a straight cutoff at 21 points, there should be a progressive scale that makes it harder to get grants as you accumulate them. One idea is to weight grant scores to the amount of effort you assign to the grant so a junior investigator putting 40% effort on their second R01 will have an easier time than the senior investigator putting 10% on his/her fifth R01. Effort levels will also have to be weighted to account for how much hard support the investigator receives from his/her institution, so the value that matters is the percent effort of what’s not covered by your institution.

  190. I agree that this action is well-intended, but nonetheless may be based upon flawed logic. I have two reactions so far:

    1) It seems counterproductive to limit PI’s to 3 R01 grants – or worse, as Marisa has pointed out, to a combination of 2 R01s and a T32, or 2 R01s and a small sub-contract on another PI R01, etc. How many of us, if faced with a fundable score on a third R01 equivalent, would jettison our roles on T32s and other collaborative grants, leading to a negative impact on collaboration and training? I wonder if the NIH has considered these potential impacts (see point 2 below)? A 4 R01 equivalents seems a more appropriate cap that would impose less disincentive for collaboration and mentorship. Alternatively, instead of abruptly cutting off successful investigators from all NIH funding, those investigators who have reached the cap could be put in a separate pool to compete with each other for NIH funding. I believe this is similar to the separate Project Grant and Foundation Grant pools currently implemented by the Canadian Health Research Institute – maybe some of our northern colleagues can comment. Or, if the NIH is really interested in promoting research funding for young investigators (see point 2), make any additional NIH funding for a PI at the cap contingent upon partnering with an ESI, in a manner that supports ESI research but does not jeopardize ESI status.

    2) I’m pessimistic as to whether this approach will actually solve the problem that it is intended to solve – i.e., to provide more grants for young investigators. The NIH is not increasing the total funding available as part of this plan, so the pay lines for non-earmarked applications will not go up. In fact, the NIAID pay line has gone down 2 points since last year. And, the relative capacity of NIH peer review to resolve the quality of grants that score better than a ~15th percentile will not increase. So, getting a fundable score on any individual submission will still be as difficult as ever, a problem that increased ESI pay lines has not solved for most ESI investigators. I suggest that this policy is most likely to benefit those of us who are ‘next in line’ after the mega-lab PIs are eliminated from the race – namely, established mid-career investigators who are skilled grant writers and are versed in navigating study section protocol and culture.

  191. While serving on study section, I saw an application from a big lab get fantastic score recommendations when the PI asked for only 5% effort, and the three young people who had produced the preliminary data had all left the lab. The grant proposed to hire new tbn’s, and not a single experienced senior postdoc was named. Although I was not one of the reviewers, I raised the issue with the whole study section: this is wrong. One of the responses: “Well, they’lI learn from other lab members”. I whole-heartedly agree that more PI’s should be funded.
    Several commentators said something to the effect that “distributing funding is communist”. I don’t think they realize just how clubby science is, and how reviewers defer to PIs who wield some power. The NIH should establish guidelines so that big labs have to plan more realistically; there will always be some flexibility at the level of Council and Program Officers. Are you the head of a big lab? It’s time you rediscovered just how much fun science is when you do it yourself.

  192. I wish the world weren’t competitive but that’s the way it is. The proposed changes may work transiently but soon we will be right back to the same level of hypercompetitiveness. Given this, I think the best thing NIH can do is improve its ability to discern the best 10-20% of grants and make it a fair competition. Only the very best science should be funded. Young and old labs are equally capable of doing good or bad science. There may be some labs that have dwindling productivity at 3+ grants but there are many labs that do not. So rather than moving to socialism–which i think will be death to American science–I think that the decision about funding a new grants to a lab that already has 3 grants should be done at Council on a case by case basis. We did this when I served on NINDS council a few years ago and this worked very well. Some labs clearly merited the extra funding and some did not. If NIH insists on going to this new system it should also given points for serving on NIH study section–coupled with a mechanism to ensure the best scientists are asked to serve.

    1. I think that the decision about funding a new grants to a lab that already has 3 grants should be done at Council on a case by case basis. We did this when I served on NINDS council a few years ago and this worked very well. Some labs clearly merited the extra funding and some did not.

      Why should we abandon the democracy of study sections deciding which PI’s “clearly merit” the additional funding and turn it over to a much smaller group of Council members? It strikes me that this is a recipe for the elite insider club of a given I or C getting all the exceptions and the less-connected types being viewed as not clearly meritorious. It will also be fascinating the see if the exception judgments (whether done by Council or by Program staff) uses the same productivity measures that Lauer and Collins are using to advance the general policy. Somehow I suspect it will not and will instead veer off into assessments of significant work decoupled from publication numbers. Which btw, is very much a strong argument against the productivity measure being used at all.

      With respect to points for study section service, it is unclear if you are suggesting adding or subtracting them from the magic 21 threshold. From context it appears that you are suggesting that study section members should get extra grant allowances? And that somehow “the best scientists” should be selected by (?) for this special treatment? This is very poorly thought out. It suggests that if only the system could find the right people and give them all the grant money that all would be well. It betrays a fundamental misunderstanding of the greatest strength of the NIH funding system and hardens one of its greatest weaknesses. That strength is the democracy of opportunity for good scientific ideas. The weakness is the inherent conservatism of like-minded folks selecting the next like-minded folks for success in the system that selected them.

      The insanity, of course, is the unwillingness to recognize that within such a competitive system that this NIH extramural funding has become, the best scientists are created in large part by the funding (all sources of input) that has been available to them. This is not a recipe for creativity and diversity of thought in the portfolio of the NIH.

  193. I applaud the goals of the GSI, but as an early-stage investigator at a large hospital institution, I strongly disfavor this proposal. Senior investigators provide the leadership required to manage large collaborative projects and translational research centers. Their own grants (and mentorship and collaboration) often support junior investigators as they transition to independence. In its current form, the high points placed by the GSI on T32 grants, multi-PI grants, and large institutional grants would have a dramatic negative impact: contrary to the stated aims, the GSI would likely cause transitioning junior investigators to lose funding, and would discourage or hinder senior PIs from training, collaborating with, and supporting junior investigators.

    The GSI disproportionately influences those who spend nearly all effort performing extramural NIH-funded activities — for instance, support through classroom teaching, non-NIH research and clinical work are ignored by the GSI, yet consume significant amounts of an investigator’s time. I also share concerns voiced in other comments for the methodology used to claim diminishing returns. In addition to rethinking the points allocation to grants described above, alternative solutions include increasing the funding pool specifically for junior investigators and taking a more individualized approach to measuring “return on investment”.

  194. I very much support fostering young academic scientists. The 5 to 10% scoring advantage to new grantees was and remains a great innovation to the granting process. However, I strongly oppose the new proposed guidelines limiting individual PI NIH support.

    The current plan under consideration will breed mediocrity. Moreover, this plan will likely drive some of the best academic teachers and investigators into alternative careers where funding is not as restrictive. This is a real concern as several colleagues have already made such intentions clear to me.

    I would support changes to CSR that might result in self imposed limits. I venture to guess that the most well funded investigators spend relatively little time on a regular NIH study section. Is such data available? Perhaps some minimal mandatory service should be required for those who reach a certain level of NIH support. Such participation would also likely benefit the review process.

  195. It is quite amazing that once science is at the point of literally revolutionizing human health that there is such a short-sighted view from the administration. The NIH has always been the shining light, and index of success and a forum for all taxpayers within the US to share in advancing the health of our nation. This is repetitively realized when Congress (in its wisdom) increases the annual NIH budget. Advances from NIH dollars literally transform healthcare as we see it, create a boom for biotech, and it is more and more evident that increasing the NIH budget proves to be a huge investment.

    Rather, the current administration wishes to decrease science and innovation at a time of economic growth. Worse still, this NIH new policy serves to penalize those who are successful, those who serve collaborators and those who wish to train the next generation. It all seems so upside down. In what other discipline does one get penalized for being a successful government grantee.

    In all areas, this new policy and structure is flawed. Merit-based rewards are always going to best approach for funding through the NIH (or any governement type contract). Placing sanctions and limits on success are beyond what we have been taught in the USA for years.

  196. This policy is long overdue. As an young investigator, although I already obtained two R01s, I wholeheartedly support the GSI system. I see so many senior investigators in my field who have way more than 3 R01s/P01s/P50s simultaneously, but produce very mediocre science/publications. It is really not uncommon to see the total output of a $2M R01 with a a few publications in very low impact open access journals, and their studies are very sloppy and some are even not related to their proposed research. However, their lack of high-impact science and productivity does not prevent them from getting the next large grant because they have the resources, connections, grantsmanship, etc.. Although not all, but I see many of them lack the enthusiasm, desire, or capability to produce the best, innovative, and useful outcomes. Many of them even do not understand what are written in their applications because most of them are written by junior investigators in their labs. It is close to a joke to say they spend time mentoring the students and junior faculty. They do not care those issues and the thing they care most is to gather whatever funding they may access and then hire junior faculty/postdocs/students to do the real work.

    A productive investigator who is really doing amazing science can always easily find other resources to support their work. The new GSI system will definitively help prevent the precious tax dollars from being wasted by those people who are only good at getting R01s but not able to deliver.

    Nonetheless, the GSI system also has limitations like, as pointed out by many people here, the excessive points assigned to MPI, R21, R03 and training grants. Despite these minor limitations that can easily fixed, the GSI system is an important step taken by the NIH that will lead to a more transparent, fair, and efficient future of the biomedical research of the US.

      1. The GSI appears to target the number of grants that investigators have over the lifetime of their career and not just what they current have ongoing. I am not in favor of the GSI as a means of putting more research dollars in the hands of young investigators. As pointed out by many responding on this blog, good science is good science and should not be judged by whether someone has had a greater share of the available resources. If we discourage the senior investigators through the GSI mechanism, who will mentor and collaborate with the young investigators? Young investigators today need a longer period of mentoring from those more advanced in their careers (yes, and often well funded both currently and over the lifetime of their careers). With shrinking dollars, a new investigator can often promote their careers by submitting grant applications with more senior investigators with name recognition, and then when they believe they have name recognition in their own right, submit on their own. The reputation of a lab and all who are part of it does influence outcomes in study sections but that is probably how it should be. Science is self correcting and reputation does count (young or senior level). Having been a long-time reviewer, I can confidently state that study sections are overall quite fair. Sometimes you win sometimes you don’t but if you keep trying funding will be forthcoming if you do good science.

        More dollars can be put in the hands of all (young and senior) if we can remove set aside programs that take funds out of the investigator initiated resource. I continue to believe that although collaboration is overall a good goal, large scale collaborations encourage mediocrity because communication breaks down as more investigators are added. This leads to just a few doing all of the work and yet the consortium grant pays for a lot of individuals that aren’t contributing.

        Now that we have all expressed our opinions, the question is does it make any difference? Is the policy set in stone by NIH and not open to negotiation? I sincerely hope that is not the case.

        1. @Surely there is a better way: What you are saying is EXACTLY the problem. “Reputation” and “name recognition” should have nothing to do with grant award, but only the science proposed in the grant and the preliminary data. If a “reputable” scientist cannot provide science of value, there is no reason to further fund his/her research based on reputation. The fact that these investigators already have a lab and employees to support should be immaterial when evaluating a grant application. But that is clearly not the case. The fact that you have been a long time reviewer and you have this kind of view of how to evaluate a proposal is EXACTLY the problem that I hope this NIH directive is going to eliminate. In fact I would propose that the grant review should be double blinded, just like some leading journals are trying to implement in reviewing papers. Unfortunately this would not be possible since, unlike papers, grants usually have preliminary data who’s authorship is easy to recognize by the reviewers. But each grant application should be evaluated at face value, independently of the names associated with it, of the names of the mentors, of recommendation letters (which need to be eliminated), of the number of previous grants (which should be also eliminated from the evaluation). Regarding the longer period of mentorship which you say is necessary, well, it is necessary precisely because grants are not available because they are hoarded by a small number of investigators. Why give all the money to a small group of investigators, who then decide what is to be studied, and who pay the people working for them (keeping a portion of the money for themselves), instead of directly paying the people who do the actual research? If someone really requires a long period of mentorship, that person will never be a good scientist. In my case, I provided my own idea even as a PhD student, and even against the resistance of my mentor that my idea is not going to work. And I ended up with a Neuron paper, which mostly helped my mentor, who had limited contribution to it, but did not help me at all because I was not “graduating” from a famous lab. That shows that even graduate students can have great ideas, but the money still goes to their “mentors”. This is ridiculous. A post-doc should be completely independent, and the lab of a mentor should be just a venue where the post-doc can learn new techniques, test his ideas and publish independently of the mentor. This is how it used to be. The mentor should only help with discussions, and not assign projects to the members of his lab as if they were his employees. This is a diversion of the initial idea of how the system was designed to work. Clearly the points of view of the “haves” and of the “have-nots” are almost impossible to reconcile, because the “haves” do not seem to see the problem. Such “experts” can often become a drag in promoting new scientific ideas. Sure, scientific circumspection is good, but when the experts turn down an idea without sound scientific arguments, just because the author used for example regular microscopy instead of confocal microscopy, or because “everyone knows that it is the other way around” (as I’ve been told), then the “experts” become a drag and should retire.

          1. Gabriel I think you may have misunderstood what was said. I agree with you completely that the science that is in front of the reviewer is what should count. I know that I have always looked at the significance of the proposed research and the methods for achieving as foremost in my scoring of proposal under review. I believe most my grant reviewer colleagues judge on this basis though some may not. My point about name recognition is that it can’t hurt to have someone with a good reputation as a co-investigator or Co-PI. This is helpful not only to young investigators but more experienced researchers as well. Although I am not a young investigator, I can definitely say that I have benefited from having a Co-PI with better recognition in an area that I have not previously published in. I am pleased to be able to draw on that person’s talents to obtain funding for a great idea.

            I continue to believe that any sort of quota system for awarding grants is not a good idea. Discrimination on the basis of ethnicity, sex or age is just wrong! A system that penalizes those with a history of obtaining more funds will disproportionately affect those who have lived longer and are chronologically older. Good ideas come at all ages. The problem is not who is getting the grants but how can we convince the current administration that more funds are needed for NIH.

            I hope you won’t give up on submitting grants — it sounds like you are a very creative person going back to graduate school days. We have all had grants turned down, we just need to keep trying. When funds are tight the chances of success are lower no matter one’s age or reputation.

          2. The science that is in front of the reviewer is what should count the most. Most reviewers do follow this recommendation by CSR, looking at the significance of the proposed research and the methods for achieving as foremost in scoring of proposals under review. Having someone with a good reputation as a co-investigator or Co-PI can be helpful to both young and experienced investigators, however, as it assures reviewers that the expertise needed to complete the research is present.

            I continue to believe that any sort of quota system for awarding grants is not a good idea. Discrimination on the basis of ethnicity, sex or age is just wrong! A system that penalizes those with a history of obtaining more funds will disproportionately affect those who have lived longer and are chronologically older. Good ideas come at all ages. The problem is not who is getting the grants but how can we convince the current administration that more funds are needed for NIH.

            It is my hope that young investigators discouraged by not getting a particular grant will keep trying. We have all had grants turned down, but my experience is that good investigators (young or older) eventually get funded if they resubmit a sufficient number of times. When funds are tight the chances of success on a given submission are lower no matter what one’s age or reputation.

          3. So you think that a sustained record of scientific discoveries and achieving the aims of previous awards is absolutely irrelevant for the prospect of accomplishing what the grant proposes…Interesting…

  197. Many colleagues here claim that the GSI will hurt those most successful investigators and drag down productivity. I think it is important to note the following issues. First, it remains to be seen whether success in getting grants equals success to generate amazing science. Second, it is essential to point out that the saved funding will most likely be distributed to those early- or mid- stage investigators or single-funding investigators whose scores are close to payline in the range of 10-15 (using NCI as an example) but not those whose applications are not discussed. This is also a group of elite investigators. If the NIH is able to develop an index to gauge the impact of publications vs. the total amount of funding, we can compare the productivity between the two groups.

  198. While the goal to increase funding to early investigators is laudible, the 21 rule is disadvantageous to them when CoPI’s are scored as 6. Rathe,r they should be scored as 7 divided by the number of investigators.
    The bureaucracy involved may be overwhelmingly daunting, in the face of impending crises for NIH.

  199. In general I support some kind of cap strategy. However, the proposed way that points will be distributed needs serious consideration and revision. For example, why is a single PI R01 given 7 points but a MPI given 6? There should be 7 TOTAL points for a R01 that are split equally among the MPIs. So for example, for 2 MPIs, each should get 3.5 points. For 3 MPIs, each should get 2.3 points, etc. Giving only 1 less point for a MPI R01 where the work and money are split equally than for a single investigator R01 discourages team science. In addition, under the current proposal PIs of training grants get 2 points even though in many cases none of the funds actually go to the PI’s lab. This is in conflict with the current thinking that graduate students should be mostly funded by training grants rather than R01s, since again it punishes the training grant PI without a benefit to their research program.

  200. I am opposed to this proposed policy for several reasons.

    1) It is based on metrics of publication, which is only one of many metrics that defines successful research programs. Missing are other metrics such as inventions, products on the market, number of companies formed, number employees who are hired by companies resulting from research, number of patients who have received better care, etc. As some investigators mature, their focus turns towards these metrics, as they are also important ways to define the utility of research to society.

    2) As mentioned by many, this policy will be harmful to collaborative research. The high point value for MPI grants will cause investigators to not seek out additional PI collaborators. In addition to stifling collaboration, this policy will hurt junior investigators who rely on the MPI mechanism to receive grants. Similarly, the high number of points allocated to program project grants will decrease the incentive to bring groups of researchers together to undertake large, ambitious projects.

    3) The proposed plan will be harmful to translational research. Translational research requires substantial infrastructure, personnel, and resources that require multiple grants to support. The GSI policy will make it more difficult for labs to sustain significant translational research programs.

    4) The policy curbs the progress of some of the most productive medical researchers in the US. It will kill potentially revolutionary concepts from these researchers that could make a real difference in health care.

    I do understand the need to increase funding for junior and mid-career scientists. In my opinion, the GSI is just not the right solution. I would be more open to a sliding payline scale. At least then, great ideas from productive investigators in research proposals that received an oustanding score would not be thrown away but could still come to fruition for the benefit of patients.

  201. I agree that supporting young investigators is a worthy endeavor. However, I also think the proposal as conceived in its current form is flawed. Specifically, the assignment of such a high point total to each PI of a multi-PI award penalizes rather than rewards the use of the multi-PI mechanism. While young investigators are supported in getting their first R01 award by the early stage investigator system, it is often getting the second award that proves difficult. In today’s world, a single R01 is insufficient to support a lab, particularly for MD scientists, so a second grant is essential. However, it is extremely difficult to get a second, independent R01 for a young investigator. The multi-PI mechanism allows a young investigator to join forces with a more senior PI to overcome this issue. The proposed budget for a multi-PI award is frequently way less per investigator than a single R01 (ie not a double module). Given these issues, it does not make sense to assign a point total of 6 points PER INVESTIGATOR on a multi-PI award. This approach disincentivizes a practice which should instead be encouraged. I hope you will reconsider this policy.

  202. This policy is long overdue. Any type of NIH grants (R, T, P, etc.) should receive points if one serves as PI, co-PI, co-I and has % of effort (no matter how small) on it.

  203. As an early career investigator I applaud and support the NIH in this effort.

    I agree with other comments that mechanisms such as training and core grants need to be carefully considered, and in most cases scored differently (lower) than research project grants.

    I understand that this places many senior investigators in a tight spot, and unfortunately I recently witnessed a the chiding of the NIH at a major national meeting. I hope that these respected individuals will instead begin to use their clout to push their institutions to explore and incentivize more diverse sources of funding.

  204. I have three comments pertaining to specifics of implementation of this plan.

    a) The points assigned to P30 core grants are in my view too high. I am the PI on such a grant, and I regard my effort on this grant as a service to the research community. Other core grant directors I have spoken to have the same perspective. At the point where I had to choose between administering the core grant and obtaining funding for my own research program, I would have to resign as the core grant PI. I am not under the impression that discouraging service is a goal of the point system. Perhaps 2 points, as with training grant PIs, would be appropriate.

    b) I don’t understand the rationale for having MPI grant points be almost as large as for a sole PI grant. This seems like an excellent way to discourage truly collaborative projects.

    c) I have not heard much about how the timing of counting points will work. Since it often takes several rounds of review and revision before a grant is funded, continuity requires submitting applications prior to the end date of current grants. I suggest either that it is allowable to exceed the point limit for a short period of overlap time, that it be possible to delay the start of a project for some reasonable length of time so that one doesn’t exceed the limit, and that grants in no-cost extension not count towards the point total.

  205. Does anyone know whether other sources of funding will count toward total NIH points? For instance does NSF funding count? Or American Diabetes Association?
    Thanks, Ian

  206. At the end of the day we are scientists and must support our conclusions with solid data. There is ample skepticism being raised around the limited and apparently skewed data being used to justify implementation of a seismic change to the current funding system. This alone should warrant stopping any forward movement of capping the number of awards a researcher can have at this time. In fact, when I look at the very same data are plotted as a linear scale, it actually makes a pretty strong case that more money = more productivity. Maybe Drs. Lauer and Collins can use this new presentation style to instead make a strong case for more money into the system. Overall, there seems to be more consensus that changes in indirect policies maybe a better place to find more money for new grants. Let’s implement change there. The current plan by definition punishes success. Any ambitious and bright person should be leery of such an approach. So should leaders of the NIH and our country, which should be striving to maintain our leadership and competitiveness in international biomedical research.

  207. This is one of the best initiatives NIH is undertaking, and I applaud the entire team for their bravery. I am not concerned about “redistribution” and “injustice” because we will never run out of exceptional ideas, and the money that is liberated from “mega labs” will certainly go to young investigators who are scoring great at their study sections anyway. Unless the established PIs can show convincing quantitative data that these “mega labs” with 30-40 postdocs are operating as efficient as 30 smaller labs, this rule should go into effect immediately. Let us not forget that given the current budget situation, we will almost certainly have a funding deficit for young investigators even if this new policy was enforced aggressively.

    It is great that NIH is taking this feedback seriously. Removal of training grants and clinical trial networks from GSI scoring was a great move; however, I think that GSI scoring for co-PI grants should also be reconsidered. NIH should encourage collaborations, not discourage them. Under this scheme, co-PIs receiving less than half of an R01 are treated like they almost have the whole funding. It should be adjusted so that one can have two R01s and still collaborate in multiple multi-PI grants. At the end of the day, it’s all about how much can a single person produce efficiently; and a multi-PI grant by definition does not require 6/7th of the commitment of a single-PI grant.

  208. While finding ways to better fund young investigators is important, the system as currently proposed will simply inhibit good science, effective training programs, and collaborative initiatives. Firstly, 3 RO1 grants are barely sufficient to fund labs with large mouse colonies, or who are doing “big data” genomics-type research. A cap of 4 RO1s would be more reasonable. Secondly, penalizing multi-PI grants at almost the same rate as single PI grants will result in fewer collaborative studies. Thirdly, penalizing P30/P50 center grants will discourage successful PI’s from spearheading these community-minded initiatives that often don’t much help the PI’s own research. Fourthly, awarding points to T32 and other training grants will discourage accomplished PI’s from running such training programs and will therefore disadvantage our trainees.

  209. It would be hard to think of a policy that would dis-incentivize collaboration across disciplines more than this one. For investigators with a major career goal of fostering research at the interface of traditional disciplines (biology and medicine, physical sciences and biology), this policy would be highly destructive. Investigators would drop multi-investigator proposals in a heartbeat, and a modest reduction in the penalty points from 6 to 5 would not address the problem.

    1. Agreed. I think the potential damage to highly collaborative scientists is one of the biggest problems with the GSI as proposed. Those extremely collaborative people (I include myself in that bucket) come at science from a different perspective. To lose that perspective and approach will do nothing good for the enterprise.

      The other major problem with the GSI as proposed is assigning any point value at all to training grants (T series, e.g.). Why in the world would the NIH disincentivize training? If the purpose is to stem the tide of PhDs in biomed science, we could debate whether that’s a good thing to do and, if so, how best to do it. But penalizing PI’s of training grants is a terrible idea.

  210. This is one of the most destructive initiative ever considered by the NIH. Successful PIs are being offered overseas jobs. Labs are starting to scale down and fire people. Collaborations are being stopped. America will fall behind and will be very hard to recover

    1. @stopthis Actually it is a very good idea not to put all the eggs into one basket, and all the strings into one (or a few) hands. The scientists that are fired from one lab are hired by another. It is not the end of the world if large labs have to downsize. But it is the end of US science if scientists leave the field and become lawyers, doctors, etc, discouraged by the impossibility to obtain a grant, which are hoarded by a few well-connected investigators. If “successful” scientists are offered jobs overseas, they would go anyway if they wanted to go in the first place. If you want a large lab, you should start a company, and then you have many people work for you if you like that.

      1. Gabriel-you are confounding several issues leading to an incorrect cause and effect relationship. Finding cures and improving human health are at the heart of what NIH does. Spreading the grant ‘wealth” will not in and of itself improve this metric-in fact it is likely to slow it down by inhibiting leading edge research being conducted in larger labs that have resources to explore bigger ideas. Spreading the grant wealth artificially will only dilute research progress.

  211. A change in how the NIH funds the individual investigators or investigative teams is overdue. With the current system, we have come from some 25th percentile to under 10th percentile funding cutoff in just over a decade. It is clear that the national politic is not ready or able to expand the total level of scientific funding to match the current, and increasing, numbers of grant seekers. While cap on the amount of NIH funding for an individual is a bitter pill to swallow for many of us, it does offer a reasonable chance to diversify the pool of funded ideas. And that is a prerequisite to some out of a large number of the nascent ideas of today translating into breakthroughs of tomorrow, as is clear by glancing at the history of our nation’s ascendancy in biomedical research.

  212. The proposed point system will slow down innovative research as investigators with novel ideas will have to wait to finish their older projects before they can start with a novel idea. This is particularly the case for clinical and epidemiological research given that the current funding mechanism favors a model of stable funding over a longer period that is well fitting for bench lab science, which also have more flexibility to change course within their funded time. However, clinical and epidemiological research has very different needs and often higher costs due to the sizable patient and participant cohorts. Accordingly, without changing the entire funding mechanism the proposed point system will even further increase this inequity to advance the scientific productivity in different research fields. We would clearly missing out to be internationally competitive in research areas that require fast adaptation of novel technologies and methods, such as in genetics, cancer genomics, or immunotherapy.

    The point system disfavors interdisciplinary team science as grants with multiple PIs (MPIs) will give a disproportional high amount of points to MPIs, e.g. an team R01 with 3 MPIs gets 18 points while a grant with a single PI gets 7 points. Accordingly, the point system will encourage the development of research silos.

    While not a set grant review criterion reviewers generally consider existing funding in their review process raising the bar to get additional funding for well-funded scientist.

  213. The goal is to fund more early stage scientists. But what happens once they get past that stage? There will be more mid-career scientists and so then the competition for funding at that level will go way up. It seems really crazy to spend money to get a lot people in the door, but then not have funding to keep them going. It is better to weed out weak scientists at the earliest stages, so that they don’t get entrenched in a career where they are not really suited.

    1. I agree-its harsh to say but it is the truth. At the moment, there are extensive handicaps for young investigators (i.e. at NHLBI a 10% scoring advantage for 1st time grantees). Already, this is causing some problems with people who received funding at 24-25% on their first grant not being able to cut it at 14-15% on the 2nd R01.

      1. Agree, the career as a scientist also needs talent, just like NBA and other careers. Why we have to punish the superstars here????

  214. I strongly support the concept behind this idea. It is a given that a lab with 10-15 postdocs will be more productive than a lab with 1-2 post-docs, and get there faster. But it doesn’t make their ideas better. Measures of productivity should be relative to the size of the lab, and compared to other “normalized” productivity measures from labs doing similar research. They shouldn’t be assessed “absolutely” regardless of how much support you had to get these. This new system will help some, and is definitely a step in the right direction of letting smaller but productive *relative to their size* labs grow into a more optimal size for them. I applaud the concept, and the details can be tweaked in time.

    1. So, if a bigger lab can get the answer faster and possibly find new therapeutic approaches to treat disease (which could mean saving lives), that isn’t a good thing? Natalie-I appreciate your concerns regarding obtaining funding from the NIH, but the entire goal of what NIH does is to improve human health-as fast as possible right? Or should we wait until a less experienced investigator comes up with some ideas…maybe?

      1. Right because the super productive PIs employing 20 postdocs have all the ideas in the lab right? Sounds like some of your postdocs should have their own labs (and RO1s)

  215. As a junior PI, I support this initiative. The loudest negative voices appear to come from well-funded labs with multiple RO1/PPG type grants. Those of us who trained in well-funded labs are aware of the inefficiencies of the system. I believe the scoring of the multi-PI and training grants are destructive to collaboration and, well, training of the next generation of scientists, the general spirit of the Grant Support Index is something I applaud.

    1. Have you not heard them? They need 10 RO1 so they will be able to toss you a coin every once in a while

  216. I appreciate the concept behind this approach, but I think the implementation is highly flawed. Many of the reasons have been covered by others, so I’d like to add one additional thought – the intent of this limit is supposedly to help younger scientists, but the implementation is going to significantly harm many Postdoctoral scientists in the short term, particularly those in the final couple years of their research.

    Irregardless of whether you agree or disagree that the system of having a small number of PIs getting a large fraction of the funding is detrimental, we all agree that it is the current situation. Postdocs made decisions about joining labs and pursuing projects based on that system – and many of the best postdocs joined these big labs, as they were places that had significant funding that could support innovative and exciting research. Implementing this new approach with no ramp up means that those large labs now suddenly can’t renew highly productive and high scoring R01s – meaning that 3rd-5th year postdocs, who are already in mid- to late-stages of completing their postdoctoral work, suddenly have to scramble to ensure their projects can get support (not to mention possibly even their own salary). I know of several labs that are now having to dramatically cut back research budgets, and may even have to let some postdocs go, because their anticipated funding support has suddenly dropped dramatically with no warning.

    If this is to be implemented, it really should be done over a 2-3 year period, so that there is sufficient time for labs to stabilize their size (both in # of personnel and in # and scale of ongoing projects) to the new system, instead of rapid changes that yank the floor out from under many of the best young scientists.

  217. I believe that this change is a bad idea. Re-distributing resources to more investigators is a socialist idea that will appeal to some people, but I do not believe that it will serve the mission of the NIH, which is to improve health. Any such change should be fully vetted in an extremely exhaustive and independent fashion by the NAS, with a goal of determining the best way to spend NIH funds to improve health.

    If the NIH leadership is bound and determined to shove through this bad idea, several changes are essential. First, this change needs to be implemented very slowly over 7 years, allowing time for institutions and careers to adjust. Second, since socialism is apparently the underlying rationale for this change, the NIH should consider all funding (i.e., eligibility for NIH funding should take into account total institutional and extramural funding). Thus, investigators who receive huge sums of money from their institutions should not be able to apply for NIH grants. Funding from other sources, such as HHMI, Gladstone, Chan-Zuckerberg would also need to be taken into account when considering eligibility for NIH funding.

  218. The cookie cutter approach often turns into a bureaucratic failure because it is inflexible. There are some large labs that should have more than 3 RO1s and others that should not. A 4 R01 limit is more reasonable for large labs that demonstrate high productivity. It would seem best to have a flexible approach to support maximal productivity. Perhaps it should be part of the discussion during grant review.

    The proposed GSI scoring of multi-PI grants will have a chilling effect on these collaborative efforts. PIs will not want to participate in a collaborative grant that gives them 80% of an RO1 worth of points and much less than that in funding. Including training grants is misguided and will have negative consequences for young investigators. I have been the director of a T32 grant for 18 years, have never taken any salary from it, and it only occasionally supports one of my trainees. I do it because I believe it is an important service to the science community, but would step down if the proposed scoring system went into effect.

    1. This seems like the best answer.

      But doesn’t it seem likely that justification could always be found for such an exception?

      Perhaps instead of a hard limit, it should penalize the priority score. Literally, that score relates to how high a priority the grant is to fund, yes? So let the fresh R01s have a better chance at being funded than the 4th R01. An exceptional and important 4th R01 will still be funded over a mediocre 1st R01.

      The trick would be to set the penalty to also allow for the fact that a lab with 3 R01s probably has better grant writing support/experience than a 1 R01 lab.

      But other than that, would that address the issue?

  219. Productivity is highly variable across labs, regardless of size or number of grants that the PI has. Rather than arbitrarily applying the 3 R01s/PI rule, it seems like it would be useful to look more closely at the data and identify characteristics associated with PIs who are highly productive vs those that are not. Quality of publications should also be taken into account (perhaps the H-index), to control for PIs who produce numerous papers that few people cite.

    1. Well put. There are always abuses but these can be dealt with at NIH. It isn’t difficult to discern those that need 4-5 R01s to run large mouse intensive labs and publish seminal findings, from the folks that run low quality paper mills.

  220. It is interesting to see various points of view here. However, both the opponents and the supporters of this initiative beat around the bush, but do not seem to point out what is really the root of the problem: personal incomes. Investigators with many grants may or may not care about the outcome of their research, but they sure care about maintaining a personal income at the maximum level allowed by contract with their university or hospital. It seems that nobody even considers the idea that it is OK to get fewer grants and earn less money that the maximum possible. This problem derives from the unrealistic salaries offered by universities/hospitals of which they pay only a fraction, pushing investigators to earn the rest of their own salaries, and to earn money for the institution in the process. Much of that extra money ends up funding activities that have nothing to do with the NIH mission, such as sports scholarships and a variety of other scholarships, maintaining sports teams and sports grounds, buying real estate, maintaining campuses overseas, holding campus tours and a lot of PR activities and perks for both universities and hospitals, as well as bonuses for executives, just like in for profit companies. The budgets of these institutions have inflated like balloons, much of it being dedicated to non-essential activities. Universities and hospitals also compete for students/patients with all kinds of irrelevant perks, forgetting that they are non-profit institutions and should attract people through the quality of their services, not through PR activities. I can easily imagine very soon universities/hospitals offering salaries of $1 million, of which they pay one tenth or less, but stimulate investigators to reach the $1 million mark from grants. I can also see a lot of well connected investigators amassing 25-30 grants in order to reach that level of personal income, and claiming that they need that number of grants to do groundbreaking research and mentor 100 people. Does it sound far fetched? I don’t think so. If 25 years ago someone would have described today’s situation and the distribution of grants, he/she would have been considered ridiculous. This is very similar to the restaurant industry, where restaurant owners hire personnel with very little pay, expecting the customers to make up the difference through tips. The situation has gotten so bad that universities/hospitals hire new people based on their perceived ability to bring money (which is their connection network), and not based on their professional excellence. Very often there is no connection between these two talents. For example, a mediocre post-doc from a famous lab will be hired before an outstanding post-doc from an unknown lab because of the connection network of their mentor which is expected to help them land a grant. Which explains why some professors are mediocre at best and students make fun of them. As a grad student at a top 50 university, I had the disappointment to see a well regarded full professor, funded by NIH, write amino acid formulas with 3-bond Carbon atoms, among other things. When prompted about his systematic errors, he seemed unsure about what the mistakes were. How was he getting NIH grants as a biochemistry professor is unclear. This is not going to lead US science to great heights. I think it is time that universities and hospitals offer more realistic salaries, which they pay in full. The extra income from grants should not exceed 10% of the base salary, no matter how many grants one has. This will ensure that only the most dedicated and talented investigators will be interested to pursue a grant, ensuring the best results for the money NIH spends. Otherwise, the situation where the US spends the most $/capita in health care but the quality of health care is well behind many other countries, will become worse. Clearly, universities and hospitals will not willingly wean off of this easily available money, unless NIH forces them to do that by limiting grant availability per individual.

  221. The 3 R01 limit does not account AT ALL for the various costs of research between different disciplines. Doing biostatistics and mathematics requires a computer mostly as well as bit of wet lab and some cell culture costs (in general terms, not my field), whoever if you are doing transgenic animals with imaging and translate that into the clinic the budget is tight and does not allow for the alleged super big labs… This difference in underlying costs is not accounted for at all in the NIH proposal. Like giving a theoretical physicist and an experimental one the same amount of money, only one of them needs large equipment and thus more money.

    Also, this is a killer for callaborations. If a MPI grant gives the main PI 7 points but the other ones 6 (or as it is now discussed 5) points regardless of the actual money distributed – why should I collaborate, I will only be disadvantaged as the Co-PI. I am currently Co-PI on a grant that provides me 25k for one year, however under this it would count as 5/7th or 71% points of a full R01, while money-wise it is only 5%, so a 14-times inflation.

    So, NIH: IF you are doing this then please account for different research expenses required for the research in general and also adjust for the monetary or effort on the MPI grant, everything else is contra productive.

  222. I am all for limiting the number of grants awarded to each laboratory. This will allow young investigators and more “well-rounded” labs to flourish. Unfortunately, the way lab “productivity” is measured is typically by number of publications without considering other impacts. A high-publishing lab invests little effort in science communication, entrepreneurship, diversity and inclusion, and forming cross-disciplinary collaborations. More diverse labs will also produce the best, most well-rounded students and postdocs who will push the boundaries of the scientific enterprise while simultaneously making a positive impact on society. We have ignored and undervalued these “non-traditional” efforts for too long, and it is responsible, at least in part, for the “alternative facts” crisis we are in now. Let’s look forward to a new reward system that gives a fair chance to all scientists to make a mark, not just the elitist good-old-boys.

  223. As an ESI, I support initiatives that will help to distribute NIH funding more evenly, especially to smaller labs. While details may need to be ironed out, this type of thinking will be critical to continue funding the best, most creative science in the US.

  224. It seems the focus of actions is on the supply of support – creating more supply, basically. Has anyone considered methods to decrease the demand? In other words, limiting the demand of grant support by being more stringent in who can apply for grants? I know that limiting researchers is anathema to much of the NIH ethos but if the hyper-competition is really a dire situation, it might be worth considering.

  225. As a second-year postdoc, I fully support the spirit of this proposal. Although my passion is to do research, I am concerned that if I pursue an academic career, funding chances are low. Of K99 winners, probably the best-equipped new faculty to become NIH-supported researchers, over a third did not get an R01.

    On a less personal note, I strongly believe that science benefits from having diverse thinkers, rather than a concentration of resources with a few at the top.

    I do worry that the scoring system is overly simplistic, including the multi-PI penalty, the fact that non-NIH funds are not considered, and varying research coasts across fields are not considered.

  226. As a PI with a small lab funded by a single modular R01, I applaud this initiative. 1) The NIH and the American people need to decide if they want a small number of big labs in a few cities, or a big number of small labs spread all over. If things continue the way they are now (module shrinks, costs rise), small labs will be unsustainable in a just a few more years. Diversity of people, approaches, and styles will be severely reduced. Only “big cheeses” who have a talent for trendiness and spin will survive. In addition, the pipeline will run dry, and America will no longer be the world’s scientific leader next decade. 2) The most important discoveries are often surprises that were not planned in a grant proposal. So why do we give study sections the power of deciding who lives or who dies? That is what is happening. Not simply more money or less money. Survival or bust. 3) I think the idea of making each additional R01 harder to get (lower payline) makes sense. 4) Institutions need to support some salaries and provide stability. 5). Finally, this comments section is not very scientific. Please NIH: request thoughtful comments from a large number of PI’s at a wide range of lab sizes and institution types, not just the 1%-ers.

  227. While the idea of GSI might be well intended, I am wondering if it would not be more reasonable to use the quality and quantity of publications in high-impact journal by grant recipients to judge productivity rather than some arbitray cut points. The proposed system is fine if intended to slow down efficient and productive investigators while slowing down progress in breakthrough research. I believe that mechanisms are already in place to facilitate new and young investigators to secure independent funding (i.e., bonus points). My concern is shared by many of my colleagues and could make discourage some scientists in staying research career.

    1. @Luc Access to the high impact journals is controlled by the same small group of people, so-called “experts” who benefit from, and control the bulk of the grants. Therefore it is not a good metric. In my own experience, any idea that competes with the interests of these “experts” will be simply side-lined. Many of these “experts” have known each other for decades, have have had the same mentors, are going to the same conferences and review panels, and have a sense of old-boys solidarity, while keeping any outsiders or newcomers out. A better metric would be the number of citations, but even there, a paper in a high impact journal will clearly have a better visibility and credibility that one in a lower impact journal. Out of curiosity, if the exact same paper were to be published in Nature and for example Brain Research or BBRC, I am pretty sure that the Nature paper would get 10 fold more citations, even though now they are equally visible on Pubmed. Just simple human psychology.

  228. Topic: Implementing Limits on Grant Support to Strengthen the Biomedical Research Workforce
    Avi/Café Twin Discussion: If “Limits” to big & well known labs at well known universities in such well known hubs as Boston, Chicago, Sillicon Valley, and few others might make available to less well known in such as places as Fairfax, VA or Richmond, VA, include strengthening small tech biz via SBIR/STTR, it’s worth deeper thinking. Perhaps there will be more foundation money effort from big labs will be encouraged while more grant.gov money is made available via SBIR/STTR for technology development for all geography ?

  229. Scientists that have been able to obtain more than 3 R01s should be celebrated and not punished. These scientists also have already come up with a list of plans to game this new idea before it is even implemented that is being forced down our throats by NIH. This is a poorly thought out idea that will dramatically hurt collaborative research and punish the most successful, most creative scientists we have. NIH come to your senses and see that this idea is highly flawed and should not be implemented. The NIH Director should fight against stifling creativity and success instead of ways to hurt the best.

    1. I completely agree with your comment John. Unfortunately, it appears that it has already been implemented. According to the most recent post from NIH offered by Open Mike they summarize recommendations of those of us who have posted as suggesting to NIH that they will consider tweaking the policy but that it is already in place.

    2. Well there is a lot that goes into funding that is not necessarily related with being smarter, or more creative. Besides, the super creative among us should have no problem in getting money creatively (Pharma, VA, DOD, Private Sector, Tech Transfer….)

    3. @John You are right, so that is one aspect that NIH still needs to make sure that gets addressed, for example by preventing PIs to accumulate more grants through proxies such as post-docs in their labs. Therefore, PIs that are collaborators or consultants on their post-docs’ applications should get some points factored into the total to limit that possibility. Also the locations proposed for research need to be kept a tab of by NIH, to make sure that there are no multiple grants accumulated at the same location/lab, even if they are technically by different people. Generally speaking, an R01 holder should have an independent position, or at least a note from the department where they work promising an independent position, should the application be successful.

  230. Point A) If your lab is producing 10 Nature Papers a year there are certainly people in lab, that should have their own lab.

    Point B) Three RO1 worth of funds is enough to do nearly anything. If you are so magnificent get money from Pharma, DOD, VA, Private sources and run your 20 M dollar operations

    Point C) Kings of departments need > 3 RO1 so to keep institutions hostage to their caprices, if any investigator can have 2 RO1 nobody becomes so special as to bend rules and become exceptions.

    Point D) More people with power means fairer competition.

    Point E) As somebody before me said – Hoarders need to retire.

  231. Has NIH ever been serious on this issue? If decisions are determined by NIH’s Advisory Committee to the Director (ACD), which consists predominantly of senior hoarders, can’t you image why this has never been resolved for decades?

    1. Correction: The link that I posted is missing in above comment. NIH is not adopting GSI system. Futile exercise. What about all those studies cited for this announcement?

Before submitting your comment, please review our blog comment policies.

Leave a Reply to JC Cancel reply

Your email address will not be published. Required fields are marked *