255 Comments
This is a tough budget time for all federal organizations. NIH is among them, having experienced a relatively flat budget for the past seven years and facing a continuation of this pattern or perhaps even declining budgets for 2012 and beyond. As we consider how to continue to fund outstanding biomedical research during austere times, we are weighing various options, including looking closely at the way we manage NIH resources. As part of the deliberation process, we have put together information on some funding scenarios, such as limiting the number of research program grant awards per investigator, the total amount of awards per investigator, the size of awards, or the amount of salary support paid by NIH.
The biomedical research enterprise is a partnership between NIH and the extramural community, and we don’t expect to make any changes to our processes without an in-depth discussion with you. We are interested in your ideas about how best to manage in complex budgetary times. We have posted the information we gathered about the various options on the OER website. I encourage you to take a look at the data. There are even some interactive graphs where you can visualize the effect of the various options. Then let us know what you think. As always, you are welcome to comment here on the blog, and we’ve also set up an email box specifically to receive feedback on these issues, NIHResourceManagement@nih.gov.
Your feedback is going to be vital as we move forward on what could be a bumpy terrain. Hopefully, we can have smooth sailing with your creative ideas providing the wind!
I don’t know the answer to this aside from the obvious – one possibility is to cut the size of the awards, but not the number of people receiving them. Another is to renegotiate indirect costs with institutions. I don’t understand how so much money goes into indirect costs, and how there can be so much variation in these costs between institutions even within the same city. Is there a way for the NIH to pressure institutions to cut back on administrative costs that is not at the expense of research?
Provided that indirect cost negotiation is not possible, another thought is to award funding to academic DEPARTMENTS instead of individuals. This would substantially reduce time spent writing grants, reduce the number of grants, and decisions would be left to department chairs to allocate funding among members of their department.
Many faculty get upset about indirect costs and believe it goes to support administration. The fact is the administrative portion of indirect costs is already capped at 26%. The majority of your indirect costs support space, heat , light and power. The Institution I work at supports research with millions of its own monies in addition to Federal monies. If indirects were to be taken away- that support would be reduced because the bills still need to be paid. Cutting indirects from grants in the long run would hurt Science and is not a realistic way to manage the budget in challenging times. Without the support for the infrastructure there will be very few institutions that will perform research in the future.
Limiting the number of awards per PI seems unfair. So we limit a good scientist and award dollars to what could be less promising science just to manage the budget and say we gave out more awards?
I understand the value of F&A to research and that it is necessary. However, for some institutions, it seems absurdly high (as much as 70%) and much more so than other institutions within the same metropolitan area (where cost of living, building expense, etc are presumed to be the same). What accounts for this disparity, and should the NIH be responsible for paying for it?
It depends on what is negotiated as a separate research rate and the composition of its base and pool costs. Normally, what I see are new buildings and the associated debt costs over a relatively lower research base = higher rate. For example, you build a research facility that gets its own rate, but the volume of research done there is quite less than the main campus rate. So you could have a main campus rate of 50%, but a new research rate of 70% for the other facility. Over time, the two will get closer as the volume goes up at the other campus. However, it is important for the institution to recoup some of its investment in the new facility especially if part of the higher F&A income will be used for debt service.
Well, if this is how disparate F&A is calculated, it is no surprise that biomedical research funding is in a non-sustainable situation. Basically, this system is rewarding universities for making fiscally irresponsible decisions to build major research buildings that they can not afford. This is also penalizing universities that are fiscally responsible since they do not have debt service on their research infrastructure.
I have always wondered how so many universities whose faculty are completely soft money are putting up so many new research buildings while we work in a sardine can at my university in regard to research space. Now I know. If biomedical research in the USA is going to survive the coming fiscal crisis, NIH has to stop rewarding unsustainable behavior by universities and research institutions.
In a nutshell, the inducement to research institutions to build or keep facilities modern (capable of doing the ground-breaking work as that evolves and gets more complex) carries with it the seeds of “perverse incentives”. Even before the “irrational exuberance” of the NIH doubling period, some institutions (e.g., some in Boston come to mind) gamed this process to have buildings or institutes with F&A rates in the 80 – 120% range. Imagine a non-modular application from a place with a 120% F&A rate: award of that grant could often fund TWO 10-module R01’s rated at a level that is not statistically significant in its difference and, I daresay most study section members would say, not meaningfully different (i.e., how strong is the predictive power of the panel really to select for “the highest likelihood of overall impact on the field”?).
In a 10 Sept 2010 Science editorial (329:1257), Bruce Alberts suggested phasing in a requirement that institutions pay 50% of PI salaries as a means of slowing the incentive to overbuild. His comments are well worth (re)reading. It could significantly change the way US univerisities do research, but maybe that’s a good thing.
If you think F&A are run away costs, then one strategy is to use Total Costs more in funding decisions. When a $1.5M application with a great review gets skipped, at some point the applicant institution will get the picture.
I know of several institutions with indirects upto 80 percent ! Why should the NIH pay for building maintenance and electricity bills ? This should be the responsibility of the univ or institution. NIH should be only funding research with a cap on indirects. The Canadian system has no indirects on govt funded grants and about 8 percent on industry supported grants. I think this will free up a lot of money.
I know my grants go through 4 people (administrative) before it gets submitted to NIH. Does it require that many people? I could do with 1 person doing it.
A combination of strategies should free up a lot of money.
And how do professors in theater, arts, English, social studies, etc. usually pay for their space, heat, light and power? The university pays for them. Many universities have immense, untouched endowments, tuition increases of 511% on average the last decade, huge merchandising and sports royalties, monies from the clinical side… yet these $$$ shouldn’t support researchers?
How do faculty in the arts, etc. support themselves?? They TEACH. Many of us scientists are employed in teaching hospitals/academic research centers and do not teach in the conventional sense. We are on 100% soft money because if we weren’t, we wouldn’t have a job. It is part of the sweatshop that is “academic medicine”.
When funding is near single digits, it is not about limiting funding for a “good” scientist. Realistically it becomes much more about reducing the diversity of “great” scientists contributing their research to biomedical science as a whole. The current system will promote well funded scientists to become even more well funded while those with limited funds finding it harder and harder to stay in the picture. Science and scientists benefit from diverse ideas and multiple fields of on-going research. Putting a cap on the number of R01s per investigator will help keep the current diversity of scientific research alive in these dire fiscally constrained times.
limiting R01s is a self defeating purpose on innovation. as existing research may lead to new directions and discoveries that in itself is also diversity, promotes interdisciplinary collaborations. These should be pursued and also science that is continuation but different from the new ideas. So limiting R01s is not the solution. As always grants should be based on merit and that is what should drive the system. What one should make sure is that the issue of overlap is addressed.
There are too much waste in the USA! For example, the light in the labs were never turned off, I saw the light in the whole building were all on without any body in it. Just too much waste in the USA everywhere, too much, too much !!!!!!!!!!!! Is it reasonable for the NIH to cover those bills ???????????????????????
Without looking at acuall statistical analysis, it sounds “Limiting the number of awards per PI seems unfair.” But when we look at the actual productivity per awarded grant, it is clear that PIs holding >3-4 or more grants have less productivity per grant. With less funding, why not improving efficiency by limiting the number of awards per PI.
If “productivity” = number of papers published, this might be correct. But there are many many papers that are published in order to enhance c.v.’s rather than contribute meaningfully to scientific progress. Successful PIs may be not only getting more grants because they have good ideas, but having a bigger impact with fewer but more significant papers.
productivity is difficult to define, quality vs quantity. plus with so many online journals, some better than others, it may be misleading. Peer review although not perfect, is a better judge of how important the contribution of an investigator has been.
Since the federal government awards the institutions that the applicants are associated with, the concept proposed (individuals vs. departments) is not valid. Regarding recovery of Facilities and Administrative (indirect cost) costs, institutions renegotiate continually with the understanding that the agreements will remain in place for several years. The negotiating process takes into account the size of the operation rather than the location. Small campuses should not require as high a rate as a large campus that has many buildings to maintain and administrative costs appropriate to the size of the organization. In one city there may be hundreds of universities, hospitals, colleges, etc., all different from each other in its size and offerings. Some will have medical schools, others will have only limited schools and research opportunities. Some will be private, others support by state/county/city funds.
There is no easy answer to controlling and distributing funding, but reducing recovery of indirect costs is not the best way because it undercuts the ability of the applicant organization to support its research activity.
At my institution, many people are internally promoted and never given lab space. This results in several funded PIs occupying the very same lab space and let me tell you, everyone is paying full price rent to live in the same lab space. The institution is raking-it-in. Indirect costs must be re-evaluated as well as the practice of packing faculty in for maximum $$$/lab space.
This is hogwash. As it is, universities are filled with do nothing administrative staff whose job is to make PIs life a living hell in order to show how indispensible they are. Institutions have no incentive to keep the indirect costs down. Here is a simple idea.
Grants should go to the individuals as total costs. PIs should be free to move to any NIH approved instution. In no time at all, most PIs will gravitate to low overhead institutions which will spurr comptetion among the institutions for efficiency.
Of course no fat cat adminsitrator will go for this simple solution.
NIH should shift its investment to high rewards, high impact, fast turnover research projects critical to improve human health, such as human embryonic stem cell (hESC) research. NIH, national institutes of health, not basic sciences, continues to keep its status quo to spend $B on non-human research, such as flies, worm, and mouse that have little implications for the prevention and cure of human diseases and have caused the so-called “valley of death”. NIH received $10B from American recovery act promoted by the supporters of hESC research, but none of those funds went into hESC research, instead to adult cells, such as induced pluripotent stem cell (iPS cell) that is alternative for human cloning, even created a center for it. 2 years after President Obama signed the executive order, NIH has made little progress, still at the hESC line stage, should move faster.
Sure, coordinate with every other federal agency to remove indirects entirely, or at least set a single (low) flat rate across the board. In every other industry, companies cope with the limitations of the system instead of being able to negotiate them away (indirect rates); why should federally funded science be any different? Obviously some advance warning will be required for institutions to cope with this change.
I think reducing award sizes is a terrible idea, it’s such a crock when awards get cut just under the amount where you can re-negotiate the work that can be done. I’d rather see a limitation on total salary-support percentage available across the federal system for PIs, with the goal of eliminating pure soft-money positions.
Eliminating soft money positions will shrink the pool for sure. It will do so by affecting junior scientist transition, percentage and timeline. It also works against the “diversity” being argues as an advantage to the “One R01 for all” sentiment expressed below.
Is this a good thing? Or are you pretending to yourself that cash strapped universities will magically invent hard money funds from the State allocation or their endowment income? ( in case you hadn’t noticed, those funds are also in full retreat)
“I’d rather see a limitation on total salary-support percentage available across the federal system for PIs, with the goal of eliminating pure soft-money positions.”
Agreed…good idea..
I support a cap for funding salary support but that has nothing to do with individuals existing entirely on soft money. In case you didn’t know, anyone who does research at a medical school in the U.S. is funded entirely by soft money. That’s right. We do 100% research and no teaching. Eliminating all Ph.D. scientists working within a School of Medicine is not the answer although capping those offensive M.D. salaries that exceed $200k/year is a good start. That doesn’t cap their salary, it just limits how much the NIH is willing to pay so…they might actually have to do some clinical work to earn the other $200k they’d like to continue receiving. Does that cut into their research time? Why’d they go and get an M.D.? Choose: do research or do medicine and earn $$$$$$. M.D.s certainly can’t have both anymore and nor should they. Do research, get paid like a researcher. Do medicine, get paid like a physician.
at times when we are trying to bridge basic and clinical sciences you can’t have these choices to MDs. That will totally kill the idea of bridging and translatability. MDs earn based on the skills of their profession, huge loans, and sacrifice of long training in the prime years. so whatever they are earning is their business with their employer. Many would be happy to make large money and not do any research or even help any researcher taking their time from activities that can give me more $$$. But they don’t. At least in academic institutions many will choose to directly or indirectly be involved and spend extra time beyond their clinical work and help bridge the basic and clinical work. You don’t want to alienate these individuals. The problem with the lowering the CAP is that then the institution could force them to bring extra money and do more clinical work, not good. So the CAP where it is is fine.
why eliminate total soft-money positions? Why are those bad for the science? There are many ways to skin a cat; not everyone should be working in one mold.
Simple analysis of the average effort it takes to support 100% soft money research positions indicates that these have become unsustainable, and are on the verge of extinction through attrition. NIH should recognize this fact and make a choice: either make soft money research jobs viable again, or explicitly discourage (i.e., prohibit) institutions from having them. Unless universities, medical schools and affiliated laboratories provide a minimum of 50% hard money for research positions, it is unconscionable to “employ” PI-level scientists and retain all indirect costs without providing an iota of a safety net. NIH should not be party to such exploitation. Until institutions are forced to acknowledge that they are getting something for nothing, the abuse will continue.
That in these extremely difficult times NIH doesn’t question why “there can be so much variation in these costs between institutions even within the same city” is UNACCEPTABLE. Is it so difficult to audit these institutions and make an economic study on the reasons for such disparities?. Sound alternative solutions should be presented to the scientific community and mobilize efforts to effect change for optimizing administration cost-effectiveness in favor of solidifying research. This is everybody’s business.
Thank you very much for posting these slides. These provide real data for a debate that has been raging in the community for a number of years (often without data). Clearly, from the data, the thing to do is to limit the total money available to an individual PI. Limiting this number to $1M would provide 2000 additional grants? Those are serious numbers.
Earlier, data was posted that the average impact factor and average number of papers peaked at about a funding level of about $750k. (Nature 468, 356-357 (2010))
As has been stated by many people over the years, there is a “death wall” at zero grants. If you have four grants and lose one, then there will likely be job losses and pain, but the lab won’t shut down. On the other hand, if you have one grant and lose one, you will likely lose your lab (particularly if this happens before tenure, but often even after tenure). Extinction is forever. Therefore, I don’t want to see the number of grants limited. However, limiting the largest labs in favor of more grants to smaller labs would seem a clear advantage to the system.
However, what we *really* need is a more stable system, in which a small lab can plan for a long term scientific career. It’s the instability that causes stress and the wasted time applying for all those grants. The instability creates a feedback loop where one has to apply for many grants to keep a stable lab. This lowers the funding rate, creates more work for CSR (and the reviewers on study sections), and wastes scientists’ time writing grants that will never be funded.
Why are “extinctions” any worse than the failure to provide a new scientist to have even one chance to prove herself, qaz? The “one sinecure R01 for all” plan has a flaw that the fans tend to ignore.
On the surface, your argument is right. But, what is a “good” science? I bet no one knows for sure. In fact, there are some sciences looking not that good now, but may have great impact in the future. In this regard, we should keep “more” science, instead of “good” science.
smart enough to do great science!
I think it is a very good idea to make these data and the interactive slides available to the scientific community. However, some key points deserve clarification. On slide 2, it is stated that the current way of managing is to “bottom out success rates (doing nothing but letting the system correct itself)”. I do not think that this correctly represents the situation. As Director of NIGMS for 7 1/2 years, we used a number of the degrees of freedom shown in the slides to manage success rates. For example, awards sizes were often reduced below the requested amount to increase the number of new and competing awards that could be made. We realized that these reductions had implications for the funded investigators, but in periods of constrained appropriations, these were deemed to be less problematic than further decreases in the number of awards that could be made. In addition, NIGMS has had a long-standing policy of scrutinizing potential awards to well-funded laboratories (see http://www.nigms.nih.gov/Research/Application/NAGMSCouncilGuidelines.htm), defined as laboratories have annual direct costs from all sources of over $750,000. Note that this is not a cap, but rather a process involving program staff and the advisory council to ensure that such potential awards are carefully considered with respect to alternative awards to less well-funded laboratories. Thus, some of the approaches described have already been utilized. Furthermore, we have attempted to analyze scientific output in the context of these policies (see https://loop.nigms.nih.gov/index.php/2010/09/27/measuring-the-scientific-output-and-impact-of-nigms-grants/). Some trends are indicated but there are, of course, many challenges to measuring scientific output in a meaningful way. Furthermore, as one might aniticipate, there are large variations at any given level of support. NIH and the scientific community need to work together to use the available data to develop policies that can best sustain the biomedical research enterprise in the long run.
Limiting the $ size of awards, the amount of salary paid, and/or the total amount of funding per investigator unfairly harms scientists at private universities that don’t also have state support, and scientists in areas of the USA where the cost of living is high-salaries must be higher in these areas of the country. Any such plans by NIH must be tied to the cost of living index in each area of the USA. This is a critical point, and has not recently been mentioned or openly discussed by those at NIH, to my knowledge.
Lorrie, the cap is set at $200k/year. Just where in the country is that a “pinch” on standard of living? This cap affects almost entirely physicians listed for percent effort and they should have absolutely no problems making up for the remainder of their salary doing some extra hours in the clinic. After all, that is what they trained to do….
I’ll hope to submit my own two cents’ worth separately via the email site but wanted to provide some contrary response (along with general praise) to Jeremy Berg’s apt comment. The first thing to say is that I think comments I believe were in, of all places, report language from a Congressional Appropriations committee were spot-on: (from memory, and paraphrasing): ‘further continuation of whittling away at awards or allowing pay lines and success rates to keep on falling is not a viable strategy’. In broad terms, the point has a lot of validity that NIGMS – within the “universe” of grant types that go to its portfolio – has been able to do some good things by whittling and whittling has some truth to it, but at the same time the NIGMS strategy being noted has the downside that it condemns many awards from the outset by providing less funding than is needed to meet the stated Aims or objective, cuts length of award, etc. While providing some mitigation – especially for less expensive forms of research – the quoted strategy falls squarely within what Congress – correctly I think – points out as an inadequate approach for a future in which Congress may not increase nominal dollar appropriations for NIH (let alone appropriating what would be needed, i.e., at least increases that match BRDPI / biomedical research inflation rates). If memory serves, Congress is essentially warning NIH that rethinking certain fundamental assumptions is going to be needed. Looking at the comment below brings one matter to mind where NIH and the research community need to get real about the TP influence: why pay way more for equivalent priority (i.e., no meaningful difference in percentile) science simply because the institutions are in high-cost regions and some institutions went on a gold-plated building splurge to put booster rockets on the depreciation component of their F&A costs?
The big science networks are now utilizing a substantial portion of NIH funds. Many corporations are involved in these projects and the focus is often on perpetuating the projects. The same science, or better, could be conducted at less cost by developing new organizational structures that are more inclusive.
On slide 4 it states that approximately 20% of the investigators get 50% of the funds. This is shameful. While there is no doubt that many of these 20% of investigators are doing great work, I do have serious doubts that their work is so important that they should be getting so much money that other solid proposals must go unfunded. Statistics like that are circumstantial evidence that phenomena such as name recognition still play too large a role in awarding grants. Sure, a person must do great work to develop a name in the first place, but that doesn’t mean that every project in that lab is pure gold and that every project should be continually funded over the course of many grant cycles. Because of this and other similar matters, I highly support the idea of limiting the number of grants an individual PI may have at the same time. Smaller, younger, less well known labs must fight for dear life in order to obtain the one grant that will keep them afloat, while large legacy labs have 4-6 grants each? Ridiculous. There will be no scientific leaders of tomorrow when the postdocs and new scientific leaders of today are slowly choking to death because of lack of funding.
“On slide 4 it states that approximately 20% of the investigators get 50% of the funds. This is shameful.” It’s actually worse than that… If you look at the interactive graph 5% of PIs are getting 25% of all NIH funds. #OccupyNIH?
I worry a bit about some of these numbers on aggregate funding and wonder if the OER folks who have been doing the community a big service pulling numbers together can clarify. For items such as “20% of the investigators get 50% of the funds” and “5% of PIs are getting 25% of all NIH funds”, to what extent are the data separating out PI’s of multi-component grants (where the PI may only directly benefit from 1/5th or less of the stated award) or, especially, of clinical trials or the like? The latter may unavoidably have immense budgets, but the overall pool of people who could be PI’s of such grants is necessarily limited (because there aren’t that many such awards).
I agree, I have also checked into this and found that the person in our Department who is the PI on our graduate student training grant, gets counted as obtaining a huge amount of grant money. But in reality, very little of this goes to that PI’s own laboratory. Similarly for a Center grant in another institution that I looked into. So I would like to see the numbers broken down into RO1 vs other types of awards before jumping to conclusions.
The “20% of investigators get 50% of the funds” statement applies to our research project grants, which includes some multi-component grants. If you click on the data table for slide 9, you’ll see the whole list of mechanisms that are considered research project grants. In the case of a multi-component grant only the project director was included in the analysis. Also for the later data points, which include multiple PI grants, it’s only the contact PIs that are included in the analysis.
This is important. The arguments against this generally miss the idea that scientific progress relies on diversity of ideas and approaches. How diverse are the approaches and insight if 20% hold 50% of the funding? I would argue, not as much as it should be.
The answer to this depends on how static that 20% is across intervals of time.
RockTalkStaff?
Check out slide 9 of the presentation. It shows it being static for at least 25 years.
Not by percentage, I meant by individual PI. Does one PI tend to stay in that 20% for only a couple of years? For 5? 10? ….most of the career?
Or is there enough turnover at the top that we don’t need to worry about diversity of approaches and insight?
A vital point – the big threat in the present (and likely worsening) crisis is the decline in diversity of ideas and approaches; the traditional “school of fish” [suddenly turning together to chase the same new fashionable idea / approach] or herd style tend to win out if the pool of grants gets to the level it is at.
Name recognition and the fear that authority promulgates is a huge part of getting your grants funded. I know of ex-study section chairs whose applications are going to the same study section that they chaired, being reviewed by their buddies who they were chairing on that study section. How do you think these buddies are going to review such applications? This “old boys’ club” network breeds corruption, with those part of this club getting a disproportionate number of grants (many of which are not based entirely on merit) while the silent majority suffers. Occupy CSR!
A younger colleague and I just attended both the most recent NIH regional grants conference and the NSF regional grants conference this week, and he had an interesting observation about the difference. He noted that there was much less sense of desperation at the NSF meeting. Why did more of the attendees seem desperate at the NIH meeting? I think the answer is germane to this discussion. The research models are very different. Most NSF faculty applicants are on 9 month appointments, and NSF only allows 2 months of summer salary for such PIs, over all of their NSF funding (iirc). There are not very many soft money positions in the NSF-type arena, and certainly nowhere near the number in biomedicine. The expectations in other science, technology, and engineering parts of the university are completely different from the medical school culture, where research may be the primary, if not only activity expected of a faculty member.
In some ways, it’s an argument for limiting the total salary a PI can put across his/her portfolio of NIH grants. But if the university is expected to cover the rest, then either researchers fall out of the system when the soft-money positions disappear, or university administrators will rightly want return for the salaries, with faculty expected to take up teaching duties, etc. Faculty salaries vary quite a bit, too, and medical school salaries are generally much higher than Arts & Sciences or Engineering. The argument made for the disparity often points to the increased “value” of the medical school faculty to the university, based on their aggregate ability to bring in NIH grants. Except they’re expected to cover their own salaries on grants. It’s at Catch 22 all around.Most of the changes up for discussion do not challenge this model. The question then becomes, is the medical school research model necessary for advancement in biomedical science?
Thanks for posting the data, and to Dr. Berg for his thoughtful comments.
I agree with Anonymous Consultant that NSF should be a model.
Over the last couple of years, I’ve reviewed a fair number of P grants- program projects, P50’s, etc, and the equivalents for other granting agencies. I think it’s great that the NIH (and others) are trying to reward collaboration, and if it works well, save on resources. However, I’ve noticed that most such grant apps designate groups of 3, 4 or more investigators who are used to working independently. There are always one or two researchers who are included just because they are well known, productive, etc…and I’m seeing little added value in putting these people together just for getting funding. How about this: “mini” program projects with two investigators who want funds to collaborate…not a multi-PI plan..two projects, with agreements from existing cores or other resources (and funds to use them), and a plan for what the two of them can do together that neither could do alone.
Thanks for asking for ideas!
Sounds like I might be voicing an unpopular view here, but I would support letting the system correct itself by continuing the current approach to funding applications (i.e., success rates fall). If anything, the size and duration of awards should be increased so that PIs can spend more time guiding and doing research and less time on applications and reporting. Yes, this means some laboratories will lose grant funding altogether – it is painful but the alternatives are not only painful, they also reduce the quality of U.S. biomedical research.
Many of the proposed options explicitly reduce the extent to which NIH funding is merit-based. The assumption seems to be that society owes every existing faculty member / PI funding to continue research, even if that research does not compete well with its peers. I think this is a mistake. Don’t get me wrong, I’m not saying that the existing review process yields perfectly merit-based awards but at least that is the goal, rather than the goal being to preserve all existing laboratories. This would only make sense if we want to preserve labs through a short period of austerity; for the long-term it jeopardizes the health of the whole enterprise. If the NIH funding were divided evenly among all PhD scientists, the resulting amounts would be too small for anyone to accomplish anything useful.
Some notes on the specific options proposed by NIH via the OER website link in the blog post…
− Reducing or limiting size (or length) of awards:
Please, please, no. This increases the number of grants a lab must write to support the same, good science, and many believe that the current sizes/time periods of grants are already too small to support the amount of application/reporting involved. Cutting large projects simply because they are expensive is arbitrary. If a particular large project is believed to be unproductive and not worth the money, then it should be reviewed accordingly and its score should reflect this. Making a blanket decision that large projects should all be eliminated in favor of many small projects will specifically target some of our most technologically advanced and multi-disciplinary projects and negatively impact U.S. science.
− Limiting number of awards held by an PI or the amount of funds an PI can hold:
The idea here is to re-direct money from successful scientists to less successful scientists. How does this help research quality?
− Limiting % effort of PIs supported by NIH funding
As far as I can tell, this makes the assumption PIs in soft-money positions are less productive than those who split their time with teaching responsibilities. I know of no evidence to support this and do not understand how pure-research institutions would survive (would their excess effort need to come from institution F&A?). The review process includes assessing the quality of research produced by the PI. Making a blanket decision to dictate how much of a PI’s effort should be devoted to research vs. teaching or other responsibilities seems irrelevant.
An option not proposed in the NIH’s materials would be to shift from a project-based merit system to a PI-based merit system. This would provide some long-term stability for labs (which seems to be the motivation for many commenters) without reducing the quality of research funded.
The point of most of the proposals is NOT to move funding from productive investigators to the non-productive. As anyone who does a lot of grant review knows, there is no meaningful difference in quality between the grant scored 20% and 10%, these differences in scores often ride on small differences that are probably not related to whether the science should be done. Further, the most recent reductions in paylines have even grants scoring at the 10% not getting paid. In this case the system is completely broken, peer review can not distinguish between an 8% grant and a 14% grant. In my experience, the top 1/4 of grants that I see should be paid and the projects allowed to go forward. The writing on the wall says that money will continue to get tighter, and “bang for buck” needs to be considered. As a reviewer, I see grants all of the time with 400K budgets that are essentially paying a PI 180K, a postdoc 50K and a senior tech 75K that produce 1-2 papers per year. Yes, that is reasonable for the amount of staff, but it is WAY over priced in relationship to grants with 250K per year budgets that have a PI paid 25% of salary, a tech and some grad students that publish 2-3 papers per year. Further, the grad students end up paying back the US economy greatly since they then go to high paying jobs in industry, increase the tax base, and provide skilled workers for the biotech industry. Thus, the grant’s impact is greatly multiplied, great science is done and skilled workers are produced.
The current system does not allow decisions to be made on any sort of cost benefit basis and is thus broken. Since the status quo is not sustainable, please realize that major changes are necessary and they need to not break the pipeline of skilled workers to our scientific industrial base.
Several very important suggestions have appeared, including setting uniform indirect cost rates at least geographically and limiting the total grant dollars an investigator can acquire (taking into account not the total face value of a grant but the actual amount attributed to a specific PI). Another factor to consider is that many state institutions ask their faculty to request salary reimbursement from Federal grants even though the state budget includes their salaries at 100%. Such state institutions then feed the excess money back to the investigators, in effect as unreviewed grant supplements. This practice should be stopped. I also agree that the NIH is currently enforcing non-existent distinctions among grants in the top 25%. Anyone who has ever served on study section knows that the top scores go to people who have friends or at least champions as reviewers. Having reviewers who are strongly supportive of the science isn’t good enough. The score is the score, however, and NIH has no other information to go on. So I would advocate a sliding scale system of paying. The grants in the top 5% would be paid at 100% (see comment on budgets below) and others paid at lower levels until the 25th percentile. That would allow highly meritorious projects to receive at least enough money to keep their projects alive. The current system of budgeting of NIH grants is deeply flawed. Investigators ask for whatever they think they can get away with and study sections are reluctant to make significant changes, especially in modular grants. A better system would empower study sections to decide on a budget without input from the applicant. The basis for assigning a budget would be the nature of the work (e.g., how many animals or expensive reagents are needed), its potential impact and the urgency of answering the question posed. All budgets would be modular in such a system. I would add that NIH spends far too much money on ideas generated by their program officers rather than by investigators and far too much on buildings that turn out to be white elephants (NIAID is particularly guilty in this respect). Also, the amount of money spent on intramural research has been never been seriously evaluated. Scientists from academic institutions who are asked to review intramural research programs are consistently appalled at the >$1 million budgets many NIH investigators enjoy. I disagree with the suggestion that universities should be obliged to pay a certain fraction of faculty salaries. Doing so would eliminate research track positions.
The best comment of all!! We should NOT remove merit based project review in favor of “allowing all existing faculty members / PIs funding to continue research, even if that research does not compete well with its peers.” Absolutely correct and right on! The USA has succeeded as a meritocracy for over 200 years, and NIH should not change this policy now.
I agree with this comment. But merit judged by some scores is many times faulty. Have you noticed that many times viewers of “American Idol show” vote and give the highest score to a candidate because that the candidate is the worst. In a peer review process many times the bright ideas are killed by the peers just out of scientific rivalry. Reviewers must declare no conflict of interest with the grant they are reviewing. SRO and PO’s of NIH must scrutinize that. If my friends are in the study section I will secure high score too. Is that a merit? Reviewing process need to be better to judge the real merit.
Good point, however, define merit clearly, and show the effect of that merit.
Now, if you are a PI you can have a grad student, or yourself write a grant on that.
In brief, the NIH premise in that slide – that the system will correct itself with that option – is just plain false. Again, an unpopular sentiment to voice, but if one artificially boosts the rate of awards to ESI new investigators, to achieve an annual influx of 1700 or 1800, the arithmetic as to how unsustainable that is across anything like a full career is pretty simple. One sees the effects painfully at study section these days, with the carnage as the results of the folks at their first cycle of needing renewed / extended support is even worse than that among longer-established PI’s (which itself is a problem at crisis point).
As numerous other posts have pointed out above, more money does not necessarily equal more productivity. You are also wildly optimistic (in my opinion) about the fairness of the current grant review process where name recognition, university affiliation, etc. can count for more than the actual science being proposed. In other words there is not necessarily a level playing field here. What you are proposing is basically a “free market” approach but the game is somewhat rigged. Also see below (from “Seasoned Reviewer”) about the validity of assuming you can tell the difference in any meaningful way between a 12%, 16%, or 18% grant application. The overall point is that there is a lot of good science that is not being funded due to simple lack of funds.
What you don’t seem to realize is that the vast majority of “obvious solutions” being offered here simply substitute a different set of biases for the ones currently at play. With no assurances whatsoever that science will be the beneficiary.
The current system doesn’t offer any assurances either. There is no perfect system that will ensure that only the best science will be done with absolutely no biases at play. On the other hand, decimating a generation of productive but second tier scientists in favor of a small handful of “super PIs” is not in the nation’s best interest, no matter how you look at it.
I don’t see where this is so obvious, Wodan. Particularly since we are talking, so far, about budgetary changes that fall far, far short of “decimation”.
Have you checked RePoRTER? Grants are still being funded…
The solutions by Anne C. seem to represent the current problems and not new useful approaches that are required for these very different and difficult times. The self-correcting mentality is the ease solution and primary philosophy of many of our new representatives in congress. Instead, lets do the hard work (which, fortunately, is ingrained in most scientists) and solve some of the key problems instead of this idea that the strongest (our luckiest) will survive, which doesn’t mean the best scientists. Anne states that “The assumption seems to be that society owes every existing faculty member / PI funding to continue research”. I haven’t read anyone making this assumption. This and other generalizations (exaggerations) are what diminish constructive/productive discussions, as people then need to correct such falsehoods vs. providing reasonable solutions. Funding certainly should be extended to those in the, for example, 9-13 percentile ranking (i.e., for those funded by NIAID), as the impact of this work is mostly indistinguishable from that in the 6-9 percentile. The sliding scale funding model (similar to the NHLBI model) is one of the obvious approaches that should be used by all Institutes. Bridge funding (R56) can be a life saver but is too short, lets put those funds into extending a sliding scale.
The key problem is the assumption that if you do any kind of ‘adjustments’ to the system (like capping the number of grants by a PI), the system is no longer merit based. You are shifting funds from good science to bad science. Nothing can be farther from truth. The reason is that there are extremely good proposals being not funded right now. The top layer of unfunded proposals are indistinguishable from the funded ones. By funding those grants, you are not funding bad since, nor funding “unsuccessful” people. You are funding more great science.
I agree with the last point that the way forward is to shift from a project-based to PI-based merit model. As long as you continue to do good science and are productive, chances of renewal should be very high. It cannot be a free ride once you get into the club. But some level of stability and assurance is essential if you want people to be career scientist. The current system is brutal and devastating to many senior, mid-level, and junior scientists alike.
The PI-based merit system is already built into the review process for NIH grants: “If established, have they demonstrated an ongoing record of accomplishments that have advanced their field(s)?” A problem might occur if reviewers equate a record of funding with a record of accomplishments.
The devastation that comes with loss of funding underscores the importance of institutions making investments in their faculty in terms of salary support and internally-funded grants.
Great idea, and when there are exactly 10 researchers left in America then we’ll have all the science we need, right?
Thanks for the transparency on the data and for the helpful slides.
Considering the appalling number of labs that are on the verge of extinction and the incredibly shrinking level of NIH funding, the only feasible solutions are: A) limit the number of R01 grants per investigator to a maximum of two, and B) mandatorily limit the total RPG support (including non-R01 grants, NIH postdoctoral fellowships, etc.) to $800,000. The immediate implementation of these measures would free up enough funds to keep numerous creative, productive and deserving labs alive. Big labs would continue to do very well, even with “only” 15 people instead of 20! Let’s not forget that such labs also receive substantial funding from private foundations and from post-doctoral fellowships. Too many labs have been/are being shut down and too many valid, creative scientists are being forced to leave academia in their mid-forties for an uncertain future. Let’s not replicate the “Occupy Wall Street” situation.
This is not a merit based solution, and harms scientists in areas of the country where the cost of living is high. I totally disagree with your ideas.
“merit based approach” is a meaningless term when there are many worthwhile grants not being funded due to lack of funds. No one is saying to spread money around to every PI but right now many promising PIs are shut out because they get a 14% instead of a 12%. You are knocking over a straw man and ignoring the very real problem. You also assume that the current system is completely fair (not true, can never be true for any human endeavor) and that giving more money to huge labs will lead to better science and more productivity (again not true based on the facts available).
why is this not merit based anymore? the Pi still has to go to the NIH review process. If his/her science is good, it will get funded and most-likely the person will never have any problems to get continuos funding…it will only help to have a broader base of research, since the money can now go to other PI’s, which also have to go to the same NIH review process.
Limitation in grants per PI, the dollars per grant and lower indirect costs are all good approaches and will likely avoid that the US ends up with a gap in research expertise due to a lost generation of reserachers.
“If his/her science is good, it will get funded and most-likely the person will never have any problems to get continuos funding…” if this statement were not so tragically clueless, it would be downright hilarious.
A big problem with a limit of two is that NIH counts no-cost extended grants as “active”
I am surprised that the cost of peer review has not been considered in these proposals to deal with the declining success rate as a consequence of a flat NIH budget.
What research has been done to compare the merits of in person peer review versus write in peer reviews or even video conferencing? Are their not funding organisations who use write in reviews or video conferencing? Could the success of their funded projects not be examined and compared with those from the NIH? What is the actual cost of transporting scientists across the country for these meetings? Surely with the ready access of video conferencing, ipads and such communication devices this is extremely feasible? Plus the success of a grant application currently lies in everyone on the review panel liking it and the tone of the conversation can irrevocably turn on the remark of one reviewer irrespective of how much impact that point could have to the ultimate success of the grant. What are the readers’ opinions?
What is the cost of peer review? A good point…we don’t hear about this at all. I agree that video conferencing or skype should work well these days.
A number of NIH review panels work solely through teleconference or skype-type web applications as of a few years ago.
This was covered several years ago under Zerhouni. The total cost of the peer review system is a tiny fraction of a percent of the total budget. Well worth the cost IMHO.
I think the online, forum type review really stifles meaningful discussion. This is a disservice to the NIH’s attempts to rank proposals, IMO.
All systems of review, btw, are going to have strengths and weaknesses. Also there will be an apparent bias for some types of proposals/discussions/issues over others. Focusing on the alleged situations where one reviewer supposedly single-handedly torpedoes a grant is fine only if equal time is devoted to those cases where a reviewer saves a deserving grant. Or when that one reviewer *appropriately* torpedoes a grant application. All of these scenarios do occur.
I believe that NSF is planning a head-to-head comparison of in-person vs. video-based review panels. One additional advantage of video-enabled peer review meetings (same time, different places) is that this is a more family-friendly option.
Limiting the total amount of money that a PI can receive is probably the most effective and most reasonable way to improve success rate. Plus, there are way too many non-productive program projects and contracts that should be cut as well. NIH also has a lot of leftover money at the end of each fiscal year that eventually goes to those who have big program projects. It is not only unfair, but also non-productive because giving additional $250k to a lab that already has a 3 million dollars annual direct cost will only result in wasteful spending.
How about cutting the intramural NIH budget in Bethesda? We never hear a discussion about this solution, and there is much more waste at the NIH Bethesda campus than elsewhere in the USA ( the extramural budget component of NIH). The percentage of the total NIH budget dollars spent on the Bethesda campus should be examined critically-most scientists don’t know that it is very high relative to the scientific output in terms of papers published, etc. from the NIH Bethesda intramural scientists.
I’ve reviewed NIH intramural labs and the science is good but not at the price of more than $3 billion for the approximately 1,200 intramural PI’s. Data from NIH RePORTER shows that the median intramural project gets more than $500,000 per year, and that most NIH intramural investigators have multiple projects (i.e. lab budgets greater than $1 million are common).
How do they count F&A within intramural labs and on RePORTER? Is there an overhead rate?
That’s a good question about how intramural overhead is handled. I don’t know, but intramural core facilities show up as separate projects in RePORTER, and there is a $29,816,280 NCI project for “space and facilities management.” But even if intramural PI projects include overhead, the budgets are still pretty generous compared to most extramural grants.
This section from the PDF document “NIH Intramural Research at the Threshold of a New Era” indicates that indirect charges are generally not charged directly to PI budgets.
“How intramural programs budget centrally: The intramural programs pay for items centrally based on their needs. If the intramural program has a very large animal program, it may be paid centrally, but other intramural programs may charge their PIs directly for animal costs. Renovations are generally costly and controlled by the Scientific Director and paid out of central funds. Rent charges, Clinical Center tap, and other ORS [Office of Research Services] charges are generally paid centrally. Most of these are considered indirect charges and not charged directly to PI budgets.”
A valid point but also consider (1) nearly all intramural labs in Bethesda rely solely on expensive post-doctoral fellows or staff scientists (there are very few “cheap” graduate students and you can’t just stick a graduate student on TA for a semester to stretch dollars) and (2) the D.C. area is one of the highest cost-of-living areas in the country which makes post-docs, etc. more expensive to hire. I’m not saying their budget should not be examined but the Bethesda campus is not the simple equivalent of a university.
Depends how it is done, no? A ground-breaking clinical trial can have a direct-cost component in the millions and presumably has a Principal Investigator, but ultimately the $30+ billion need to lead to advances in human health. And many (most) of the most important clinical trials funded by NIH are work that Pharma (or the medical device manufacturer, or the surgeons pushing a procedure) would never want to fund.
I think that NIH has to re-examine the assumption behind the establishment of current system. I remember that, only when the payline is higher than certain number (30%?), the current system is able to identify the meritorious grant and fund it. I am talking about payline, but not the so-called “successful rate”, which is artificially inflated. Now, the payline is much much lower than the one that was based to establish the current system. Since we are all professional scientists, would it be shameful that we are even debating whether or not the current system needs to be changed, considering that the very foundation of the current system had been long gone? Once we understand this problem, the solution is actually very simple: Making every effort to restore the targeted payline!
One suggestion is to use tiered system. For major grants, take R01 for example. For an individual investigator, the payline of 1st R01 or equivalent may set as 30%; 2nd as 20%; 3rd as 10%. etc. The logic behind is self-evident. A lab with 2 R01 has much more advantage to compete for 3nd grant than the lab with 0 or 1 R01.
Another suggestion is to use reduced Ro1 to fund the grants outside the payline but having sufficient merit. For example, a 2-year R01, or 4 year R01 with $150,000 each year should be used to support these grants. There are many good reasons. For example, reviewers sometime criticize the validity of the 3rd aim, but have high enthusiasm for the first and second aim. If all three aims are independent, why can’t the NIH just fund the first two? In another scenario, the proposal may be too risky; but if proven true, it may have very high impact. Thus, giving an award for a short period of time is justified to further develop such ideas. The current bridging mechanism needs to be expanded and incorporated into the grant review system.
I think that the tiered system would be a good compromise between a hard limit on grant#/investigator (perhaps compromising quality) and trying to spreading out NIH funds, not just to support everyone, but to make sure that there is a large and diverse portfolio of research. An addition to the idea of the tiered system that has been proposed elsewhere is to decrease indirect costs awarded for the 2nd, 3rd, 4th grant, etc.
Right now, the paylines are so horrific that NIH could award 2-2.5 x more R01’s than it does with no compromise on quality. In other words, the approx 800 RPG awards “freed up” by a cap at, say, three active RPG awards would go to applications that were not funded but of indistinguishable quality.
I like the idea of funding per aim – sort of like buying an app for your iPad. Budget reductions make sense if tied to effort, and tying it to the executable Aims would be a good way to do it.
first 30%, 2nd, 15%, 3rd, 5%, 4th, 0.1% seems to be more fair
Or you could add 2% points for every R01 you already have, after the first two. If you get scored at 10% for your third R01, you are bumped to 12%, and to 14% if this is your forth.
You just have to make sure reviewers don’t score with that ‘bump’ in mind. Otherwise all the proposals in the top 5% would be third and fourth grants.
I agree with several posts on this blog and thank Dr. Rockey for providing the research community with such a forum.
In addition to the numerous suggestions to cap lab funding as the benefits above 750K are minimal; an approach that would make grant renewal in subsequent years less arduous should benefit the entire system and provide stability. For example providing some type of quinquennial review similar to the intramural system that will not be focused on preparing extremely long grant applications, but basing it more on past success and a strong yet succinct 5-year plan could be a better and more stable approach (taking an approach to invest in the people similar to HHMI and Welcome Trust).
The benefits from such an approach are numerous, but in particular this will allow the PIs to be in the lab performing experiments, training PhD students and post-docs in a hands-on manner. The PIs can assist in the development of projects that should result in greater output, instead of spending months on end in an office preparing grants. In addition, putting in place a system that essentially allows young investigators to have 1 tech, 2 postdocs, and 2 grad students and tenured individuals to have 1 more of each generates lab units that are stable and costs associated with the labs less variable. I feel this approach used by the NIH (intramural), EMBL, CR-UK and now the Francis Crick Institute in London are highly efficient and result in greater scientific output as well as exceptional mentoring by the PIs who are at the bench and able to help train the next generation of scientists.
One last downside to the mega-labs (besides postdoc and grad students never really receive any mentoring from the PI as they are locked up in an office writing grants and papers) is that it reduces diversity in the scientific community. Big name PIs with large labs have a lot of influence to get papers published in certain journals and then postdocs from these labs often go on to get the better faculty position, but in many instances are just clones of the mentor and their approaches to address scientific problems and develop hypotheses is very similar to the PI. However if there are many smaller lab units with different styles and approaches to developing novel hypotheses this should in theory increase the innovation, creativity, and in general scientific approach to tackle a problem.
This solution is not based on the merits of the science and is, in my view, a poor solution.
Often in a larger lab the PI does spend lots of time mentoring-why is there the assumption among these comment writers that this doesn’t occur? Also, postdocs and students learn so much from each other too, not just from the PI. In my experience, a grad student or postdoc in a small lab is much more isolated, learns fewer different techniques, and is less well equipped for the future than students and postdocs from larger labs. I would assert that larger labs are better environments for training students and postdocs than smaller labs. Thus, limitations on total support are not reasonable or even desirable.
I think that large labs that are well managed (ie, the PI expects cooperation instead of competition) are great experiences for postdocs and grad students. They learn from each other, learn the value of collaboration, and form a network of trusted colleagues that will persist throughout their careers.
Lorrie- A closer perusal of these comments posted here should convince you that merit is not the issue; there are large numbers of meritorious proposals not getting funded under the current winner-take-all system.
Why winner-take-all? Because the best funded institutions claim they need new research space = more indirect costs = ability to fill new space by recruiting those PIs who have multiple grants that can pay their total salary …. which gets us back to big medical centers building more research space. Stop the med school building sprees by limiting $$ per PI and limiting % salary/PI – even if it means the faculty scientists might have to justify their professorial title by a little teaching.
I think the major problem spurring the need for multiple grants is the need to cover salaries, and an HHMI-like approach like you outline could help in that regard.
Getting rid of the “big science” spore grants, U grants, P50s etc will free up a lot of money for individual R01s. These big science projects haves less impact and much less accountability for the science produced and the number of papers published per dollar granted to these mammoth projects compared to RO1 grants. PIs of RO1 grants are held highly accountable for productivity over the 4-5 year grant period while this isn’t the case with these huge grant programs. Also, grants for resource building should be highly limited and kept to a minimum- if individual PIs cannot sustain their own labs through RO1 grants, there will be no one left to use these resources!!!
The CTSA also looks to be major program with minimal output. For successful translation one need something to translate, and this is provided by the R01 programs that are now confined to the top 7th percentile.
I’m not convinced of the effectiveness of CTSA initiatives other than their effectiveness in siphoning money away from basic science. Our institution is hurting because suddenly basic dollar is being redirected to feed our CTSA initiative. The infrastructure to chase these grants is enormous. World class basic science programs in need of infrastructural investment languish while increasingly more resources are concentrated into the hands of the rich….um, this reminds me of something…
I think the statements about the CTSA’s reflect a lack of understanding of what the CTSA program is all about. These are infrastructure grants to enable clinical and translational research, so the “output” is actually that of the sum total of everyone who was supported by that infrastructure, and that is certainly not “minimal.” For example, the “Clinical Service Cores” that many CTSA’s have are hospital beds or outpatient sites for testing clinical research subjects. The number of high-profile clinical studies that have depended on these units and that have changed the way medicine is practiced is huge, and in fact these units alone have saved the Institutes and Centers tens to hundreds of millions of dollars that would otherwise need to be budgeted in grants for hospital beds and rooms. Don’t get me wrong, I fully agree that the pipeline that fuels many or most clinical advances is basic science, and it needs continued high level support. At the same time, however, we also need to remember that the pipeline that takes those discoveries to the clinic is equally important and is greatly dependent on the CTSA’s. The latter is, after all, the ultimate goal of the NIH (H as in Health).
My comments relate to observation on the ground at our institution, which has decided to resource less viable clinical research units at the expense of basic science programs that are in the top 5% in terms of successful funding for the institution. Clinical research is valuable, but I can’t name many Nobels for translational research, while I can recite several whose importance had great translational impact and signficance to public health. Others may dissagree.
You cite the clinical service cores as being a major benefit of the CTSA program, but the history is important here. The CTSA replaced the General Clinical Research Center program, which was entirely focused on the clinical research services you describe. The other 6-10 “key function” areas were added to the GCRC model. The idea was to integrate bench, clinical, informatics, etc. to accelerate the speed of translation. It’s a nice idea, based in large part on the success of NCI’s Cancer Center program, but what works within a discipline/disease has not translated into effective broader support structures for translational research. In my opinion, the ROI for the CTSA program hasn’t lived up to its promise (hype?), nor met the goal of transforming institutional cultures around translational research. The GCRCs had developed some of the problems that long-running programs develop, starting with a sense of entitlement in the grantee institutions, but the CTSA structure may have undercut some of the value of the GCRC-like key function areas.
Where are you getting your data for these misguided statements on impact and accountability?
It seems to me that a combination of the suggested options will work best to keep science alive.
First, cap indirect costs on grants to no more than 40% with a better justification for even that. It is time for universities to stop milking the cash cow. It is remarkable how little return the actual science gets from this money.
Second, reduce the salary cap as suggested and start to phase out NIH support of pure soft money positions. Two hundred thousand dollars a year salaries are not sustainable with the current budget climate and are extremely hard to justify to congress critters. It is interesting to watch faculty recruiting these days where universities will “offer” salaries in this range to soft money faculty while a median full professor salary for someone on hard money is probably in the range of 130k or so. This has little to do with the physician-scientist scenario, in that case most of the salary is coming from clinical work anyway.
Third, by all means cap the yearly support from grants to 1,000,000 or even 900,000 per year. While one can argue that these most productive investigators should be supported (and they should), anyone who has ever worked in such a large lab will tell you that the large amount of available resources leads to their waste. While there is still alot of productively, the “cost per unit science” is often much higher than the most productive smaller labs.
Fourth, allow grant reviewers to incorporate “bang for the buck” into the final assignment of priority score. I recently reviewed a 800,000 per year proposal that would really cost that to complete due to the expense of the technique. However, the information obtained was probably only really worth 200,000 per year. However, I was told that I had to give my priority score without consideration to the budget. NIH is the only organization I can think of that makes funding decisions without regard to cost benefit analysis. It is not that surprising that the funding rate for large grants is higher since if a grant has twice the budget, of course the PI can get more done.
Same thing for competing renewals, the reviewers should be told what the amount of the prior award was up front and part of the priority score of the renewal should explicitly be driven in part by a cost/benefit analysis of the prior award instead of a nebulous review of “productivity”. Grants in the 200K range should not be directly compared to 800K behemoths, the expected productivity should be scaled.
I don’t want to derail the conversation, and you make a lot of interesting and thoughtful points, but I have to take issue with this comment on F&A.
It is remarkable how little return the actual science gets from this money.
Most of what supports the actual science isn’t obvious, including the need to deal with federal mandates around receipt of federal funds, in addition to the usual arguments of F&A “keeping the lights on and the floors swept”. And regardless of what kind of research administrators you have (help or hurdle), they’re necessary for keeping the university (and you) within the federal rules. I’ve counseled smaller companies not to apply for federal grants because they don’t have the infrastructure to deal with the “cost of ownership”. When they don’t already have it, they can’t pre-negotiate sufficient overhead to pay for it should the grant be funded. Most researchers don’t think about the fact that when the federal rules aren’t satisfied, the grant isn’t paid. You the researcher don’t have to take the time and effort to deal with all the “OMB circular #123” issues when the research administrators and OSP accountants, all part of F&A, do their jobs.
For another example, my PhD lab was in an urban, high crime setting, and campus security systems and police made a big difference in my daily life and ability to do “actual science”. This kind of thing becomes invisible to us, but F&A pays for it. Oh, and when I helped set up a lab for a start-up, we had to figure out how to safely dispose of the ethidium bromide, and then I really appreciated what F&A paid for.
This does not explain why NIH pays two different indirect cost rates (32% versus 50%) to two different colleges at my comprehensive university which makes it a big pain to submit a dual PI proposal with PIs appointed in the separate colleges. The campuses abut each other (and the campus centers are less than 1 mile apart). The NIH funded science at the two colleges is similar in scope and there is no apparent difference in costs of doing business (all of the research infrastructure such as safety, research administration and core facilities are shared throughout the university and it is not like one needs more police than the other). I am not arguing that indirect costs are necessary for support infrastructure, what I am arguing is that the mechanism by which rates are calculated is arbitrary. I would propose an equal indirect cost rate across the board, or at least a standard one for urban versus rural institutions if the need for difference is cost of land and policing. One university should not be getting 70%+ in indirect costs while another gets 30% or so.
While these comments seem closest to the mark in many ways, I think at least two points of nuance are needed. (1) An F&A rate of 40% is probably unrealistically low; 26% is the admin cost component for educational institutions* and has long been known to be below actual costs. (2) The assertion that the physician-scientist scenario is not a concern because “in that case most of the [$$$] are coming from clinical work anyway” is between dubious and ludicrous. I tend to agree that NIH / Congress need in thoughtful ways to reduce the salary cap for non-clinical projects but there would have to be some nuancing or tiering based on clinical activity unless someone wants to amplify an already dangerous trend (the declining pool of research-intensive clinicians for the portion of the NIH mission that involves clinical research, trials, etc.) [aside – I’m not among those blessed with salary cap issues or clinical duties, and yes, I think institutions are going to have to cost-share the salary component better but, realistically, the idea that it will help sustain public support for NIH & its funding of basic science to chase even more clinician researchers off does not strike me as a good one).
* I heard once, and have not tracked down clear evidence that the assertion is correct, that free-standing research institutes (non-educational institutions) are NOT subject to the OMB-imposed 26% cap on administrative component of F&A (indirect) costs.
If true, one ‘small’ item (probably worth dozens of extra R01 per year) would be to extend this cap to ALL NIH funding / F&A, not just that for educational institutions. That should be near the top of the “must-do” list [despite the firestorm of push-back from influential / powerful institutes].
Looking at other industries, you see things like health-insurance companies with a legal requirement that they must redistribute 80–85% of their income; that means they not only survive but profit with 15–20%. I don’t see how we can justify anything more in science, when you consider that health insurance is one of the most profitable industries right now.
Huh?
F&A rate is not profit.
Educational institutions at which NIH research is conducted almost all “lose money” on every grant they get, that is, they almost all fall well short of the “full cost recovery” that ostensibly Congress mandates. The OMB-dictated cap on administrative cost component of indirects (26% of modified direct cost base; if memory serves, that was a number David Stockman chose – probably somewhat arbitrarily – in Reagan era as OMB Director) is lower than what universities ever can get to, especially as the administrative burden gets ever larger. Short form, universities need to cost share on every grant. If it has not done so, OMB ought to extend the same cap to free-standing, single-purpose research institutes and any other entities getting NIH awards. Every bit of money freed up will help in the current crisis and coming storm.
Mark B-
So are you suggesting the free standing research institutes should be driven out of business? Are you asserting (on what basis?) that they are just scamming the Indirect negotiation process and pocketing huge sums of pure skimola? Or do you think there is a secret source of funds they will tap once the Federal tap dries up?
well, well. Speaking of the salary cap, your CongressCritters are listening
There is no single answer to managing NIH resources to promote a broader distribution of funds among investigators. It is the responsibility of each institute to manage their portfolio in a responsible manner that targets a proportion of the portfolio to predetermined areas of interest or of perceived need (by congress and/or the research community), while setting aside proportions to foster new and young investigators and for work that will continue to move a field forward by building on the previous work of others. In some cases the work will be of high risk, others will be less so, and finally there will be cases of more plodding studies (i.e., work that is more descriptive).
The institutes have progressively been placed in the position of taking a less active role in directly managing their portfolios to these ends, instead relying on an inept review system. Reviews composed of nothing but terse bullet points offer little constructive feedback to the institute (let alone to the PI) to judge both the value of the application or of the quality of the review. Hence, funding decisions are based almost totally on scores with little information to help NIH staff and council members more clearly assess what is presented to them by study sections. They are thus handcuffed into taking less of an active role in the funding process.
In the distant past, study sections were composed of actively funded investigators, a significant proportion of which were either deans or department heads. The Chair of study sections actually helped to guide the discussions to promote a fair review that focused on the importance of the proposed work and/or the unique attributes an investigator brings to the table. Due to the combined expertise and experience of the panel, reviewers were much less likely to get away with biased and unfounded comments that can sway the opinions of those members of the panel with less expertise in a specific research area. Scientific review officers were also more knowledgeable of the areas of research under discussion and played an active role in seeing that reviewers actively discussed an application leading to as much consensus as possible as to its strong and weak points. Yes, critiques were lengthy, requiring substantial efforts by reviewers, but they provided a much clearer framework for evaluation of the application by the PI, and just as importantly, by the institute. The system wasn’t perfect, but it was certainly more fair, constructive and informative than today’s system in which the Chair and Review Officer do little more than introduce each application and the reviewers produce little discussion or reasoning for their scores.
As I mentioned above, there is no one answer or magic bullet to managing the NIH budget to better distribute funds among investigators. It will take at least two major steps. These steps make the assumption that there are many meritorious applications that are currently unfunded.
First, all of the options listed in the pdf file are viable and should be used by each institute to achieve its targeted goal of the number of applications to fund. Although, the concept of increased limitations on PI salaries will likely incur too much resistance and is probably not worth the fight, especially if all of the other options come into play. Towards this end the institutes will likely need to reassess their administrative structure, consider increasing input from the extramural community as well as taking a closer and more thorough look at investigator progress in non-continuing renewals. Also, each institute must increase the expertise of its staff so that it can better evaluate submitted applications to actively manage their portfolio to target a set number of applications to fund. Such changes will likely result in higher administrative costs to NIH, but is a small price to pay to increase the number of funded investigators and for better grant oversight. A more active role by members of council or ad hoc input from the extramural community in combination with NIH staff will be needed to help ensure that funds are distributed appropriately and fairly.
Second, the review process has to be improved to better ensure more fair, thorough and informative reviews of applications. This will require more knowledgable Review Officers that play a more active role in the review process. Chairs should not only be experienced investigators but also have training in how to guide discussions to provide all members of the panel with the basic information needed to make a fair decision regarding a proposal, to help reviewers come to a consensus, and to stimulate discussion. Review panels must be comprised of experienced and active or recently active investigators. NIH already recognizes this need as evidenced by the carrot of an additional year of funding. But more needs to be done. Obtaining an NIH award is not a given right, but should instead be considered a passport to joining a community. As such, each member of the community has an obligation to ensure the community remains strong by contributing to the review process. In other words, if you receive an award you are obligated to sit on a review panel at least one or more times. The concept that review panels have to maintain continuity no longer holds water, as applications are now only reviewed no more than twice and panels are composed of as nearly as many ad hoc members as permanent members. Suggestions for smaller and more focused review panels that utilize newer technologies should be given serious consideration. Finally, reviewers must provide clear, constructive and informative reviews that will assist not only the investigator but also NIH staff and members of council.
Given that the current grant application review process is badly broken, going back to the “old” review process won’t make it better. We need more funded projects, and let the results determine which are good projects. Study sections will never be very good at making that determination (at least among the credible applications; they can weed out the truly awful).
What I like to do is look at the productivity of people in my subfield of science over the past 2 decades or so. Then to ask which papers have been most influential and interesting. This analysis can continue to ask how those folks got there. With a lot of funding and data and staff is usually the answer. Bang for the buck is seductive but there may also be an activation threshold for a bang of a certain size. Taking a look at what has come before can help us predict how various changes in award might affect scientific discovery.
Whatever is done should be implemented carefully in a way that the best peer reviewed science is preserved. A significant drop in success rates or paylines followed by a sudden improvement could permanently shut down excellent to outstanding research by an investigator who happened to have a grant renewal due in the lean year while other programs ultimately benefit under improved success rates. Instituting across the board cuts to all grants in order to stabilize paylines and success rates should be a goal assuring that (even if at much less dollars) the best research is allowed to continue whether it is a noncompeting or competing renewal. Over time, the least meritorious (although probably still very good) research will be unfunded while the absolute best science will be preserved for years to come. In fact, the slow retractions is in a way accomplishing this although it is painful. If Congress enacts for example a 10% cut in NIH funding in one year, it should come first from eliminating new initiatives and then by across the board cuts to existing programs allowing new and renewing RPGs to be funded at the current success rate with paylines at a minimum of 10%.
In the past several years, since the doubling of the NIH budget, the bureaucracy at NIH has grown along with the funding. Along with this is a correlated growth in the amount of research funding that is going to programmatic or solicited research and a decrease in the percentage of funds going to investigator initiated research. There appears be increasing numbers of contracts and types of programs that sound fine when proposed by NIH administrators, but often produce lots of data that is of poor quality or is useless unless someone with funding takes advantage of that data. These data are not included in your published data but intelligent estimates suggest it has changed from>60% investigator initiated in the 70s to >60% programmatic today. I would like to see the data on projected success rate of 50%, 60% or 70% were guaranteed to investigator initiated grants, and these analyses might show there is not a funding problem just an inappropriate distribution of funds. One idea is to set the number of investigator initiated awards to ensure a certain success rate and then let program compete for the rest to ensure that only the highest quality programmatic research gets initiated. As a SS member I can report that really productive and important research is not making the funding level today and there is nothing we can do to change this based on the level of funding and the number of grants that are being reviewed.
Just a point of clarification. NIH Data Book slides show that less than 20% of RPG and R01 awards are in response to targeted announcements (not investigator-initiated). See data For RPGs at http://report.nih.gov/NIHDatabook/Charts/Default.aspx?showm=Y&chartId=25&catId=2 and for R01s at
http://report.nih.gov/NIHDatabook/Charts/Default.aspx?showm=Y&chartId=34&catId=2.
Apart from the percentage of awards, how do the dollars play out? What is the percentage of dollars spent on targeted (including contracts, centers, etc) versus investigator-initiated grants? This may be more informative.
There are some links in our latest post that help answer this. http://nexus.od.nih.gov/all/2011/10/26/there%e2%80%99s-still-time-to-join-the-discussion/
American science and American Universities are great and the system of federal funding at NIH and NSF have made that possible . The system is not perfect . The greatness is built on the success of the funding paradigm a decade or more ago when the budgets were about half of what they are now. I remember lobbying with other chairman organizations to get the NIH budget up from 7$illion to 11$Billion and then to double that . What I see is an inflation of costs per paper , per discovery. Big science has taken over since the human genome project . Yet it was the RO1 mechanism that produced so many discoveries . What NIH needs to do urgently is to
increase the number of RO1 awards . The RO1awards have proved to be efficient financially and efficient in producing discoveries. One of the greatest factors contributing to American science is the number of individual scientists all over the country who were supported by RO1s . Without more RO1 awards those individuals will not be around, and the young people they train will not be in science . We may be resorting to FEMA if NIH cannot restore the RO1s to more individual scientists .
One of the good ways to train young scientists is to let them have smaller grants like R21 and R03.
May be I missed something but the solution looks obvious. Instead of reducing grant size it is better to reduce number of grants per PI. According to recent Nature publication it can be as many as 32 (!) NIH grant per one PI. And from any of the grant NIH pays indirect costs. Why? Usually research teams don’t spend more money for electricity or heating, they stay in the same space, etc.
Unfortunately, expanded indirect costs is main problem of NIH budget. Its were inflated during previous “good times” and now should be reduced leaving more to scientists.
Here’s something I haven’t seen mentioned explicitly: RFAs versus unsolicited R01s. Like most investigators reading this, every week I dutifully read the RFA emails from the NIH. What I find mostly are requests for big science initiatives that are of questionable benefit for the cost. It sometimes appears that somebody making programmatic decisions has decided to promote a pet project. Now, I don’t claim to have a good view of the research portfolio landscape, but here are some examples of RFAs from recent emails that I would judge to be of dubious cost vs benefit to society. Let’s fund fewer (or at least smarter) RFAs and more unsolicited R01s. I don’t have specific vendettas against the RFAs below, but I do question who is making the cost/benefit decisions that result in pre-allocation of funds and thus take away from the majority of investigators whose research does not qualify for the RFA.
Economic Studies Ancillary to Completed or Ongoing Health Care Delivery and Financing Pilots, Demonstrations, and Other Experiments (R01)
(RFA-RM-11-023)
NIH Roadmap Initiatives
Oceans, Great Lakes and Human Health (R01)
(RFA-ES-11-013)
National Institute of Environmental Health Sciences
Silvio O. Conte Digestive Diseases Research Core Centers (P30)
(RFA-DK-11-022)
National Institute of Diabetes and Digestive and Kidney Diseases
NIDDK Interconnectivity Network Coordinating Unit (U24)
(RFA-DK-11-030)
National Institute of Diabetes and Digestive and Kidney Diseases
Application Receipt Date(s): March 15, 2012
Institute a maximum number of grants (3), and total grant dollars ($750K/year). Any researcher can only effectively oversee a modest number of advanced projects (2 or 3), the associated students, and a certain size budget simultaneously. We all know PI’s trying to manage so many projects that their outcomes, mentoring, and findings suffer.
Reduce, or at least freeze the NIH salary cap-If for no other reason than to send a message to the public that most of us are not getting rich from federal grants.
Create a standard F&A rate (say 45 or 50%), that decreases as an institution garners more grant funding after a certain base level (maybe $50 million). If F&A is to support the administrative and physical infrastructure this makes sense. There should be an “economy of scale” in which the second or third $100 million should not need as much additional infrastructure or management personnel. A standard F&A rate would also save millions in man hours spent on cost proposals and eliminate the “gaming” of the system these proposals have encouraged. After all, in what other system do you “win” by being the least efficient institution with the highest rate? With this issue alone billions more would be available for grants.
Create a truly blind proposal process by eliminating names and institutional affiliations from all proposals during the review process. Everyone should compete on an equal playing field based on measurable research outcomes or pilot data. Those of us with experience should be able to write better proposals with more depth without having to use institutional reputation or “name dropping” to receive preferential treatment. Also, assure that all review panels have adequate representation from different institutional types and not just those with the “right” pedigrees.
Having written a fair number of proposals and reviewed quite a number as well, I am always fascinated by these proposals for blind review. How is this supposed to work? Systematically, I mean. All it does is ensure that the most famous labs are not blinded where the lesser known ones have a chance to be blind-reviewed. Not sure this is the slant the blinding fans are after, eh?
Also, there will be an attempt at gaming wherein the PI drops broader hints about her identity when she chooses, conceals it when desired and maybe tries to imply it is another competitor for other situations. No matter what rules you establish, they can be gamed.
Hey, all of you struggling PIs…… Let’s start an OCCUPY MED SCHOOLS movement!
It’s the lowest-hanging fruit:
One of the largest unnecessary drains on NIH funds= EXCESSIVE INDIRECT COST RETURN. There is no legitimate reason for indirect cost return to exceed 60%. The accounting tricks that institutions use to inflate their indirect cost return should be eliminated.
The medical school model of parasitizing off of the NIH for salaries is broken. By changing IDC policies, NIH can provide some “selective pressure” that will force them to evolve.
One such selective pressure could be:
Among grants with the similar priority score, only those with lower inderct are funded.
A big part of the problem is that the current system of “generous” indirect rates (whatever else you say about them, there is little doubt that they are generous) has encouraged universitites, particularly the most research-intensive medical schools, to greatly expand in research capacity by building new buildings, in the hope that they will be able to attract faculty who will then be able to pay rent on the new space through their F&A rates. The problems is that they cannot all be winners, and it has created a situation where a large fraction of the faculty are supported primarily by soft money and thus have an incentive to write a large number of grants, many of limited originality, in the hope that at least some of the rolls of the dice will be winners. This is not the most efficient way to run the scientific enterprise. It leads to NIH study sections making de-facto decisions affecting scientific careers that are best made by promotions and tenure committees.
What is the solution? Probably will have to include multiple approaches, but reducing the overall F&A rates have to be part of the solution. That will slow down the frenzy of construction by academic medical centers, and while it will result in a marginally smaller scientific enterprise, it will be one in which faculty are spending most of their time doing research, rather than writing grants.
Decide which research projects to support, and then fund them at an appropriate level. Cover the full costs of NIH research, both direct and indirect, consistently and across all programs.
Federal research agencies often have ambitions that exceed their budgets. Program managers routinely approve more research projects than can be funded, then provide each project only a fraction of the budget needed to carry out the research. They do this in the hope that the universities will make up the rest in the form of voluntary sharing of direct costs or partial exemption from indirect costs.
Compared to public universities, the NIH is not in a tough budget period, sadly. The agency must not solve its problem by shifting even more costs to the universities, either through indirect cost recovery or increased cost-sharing.
Universities are already supporting research at a level of $11 billion nationwide, up from $1.4 billion 25 years ago (see the NSF-published research expenditure tables.). The university-funded proportion of academic research expenditures covered by institutional funds has almost doubled over that period. Universities should allocate their resources to fulfill their combined missions of education, research and public surface. They should not and cannot support a steadily increasing share of federally funded research, as they have been.
Public universities have been hit particularly hard by the effects in recent years to move research costs from agencies to the universities. State funds have decreased dramatically. Universities are already subsidizing NIH research from these reduced university funds because the presently assigned indirect cost rates represents about 3/4 of full reimbursement. Over the last two decades, the government has added a steady stream of new compliance and reporting requirements, driving the actual administrative costs well over the cap on the reimbursement of those costs.
So, if the Universities are losing money on NIH grants and the most productive NIH grantees (3+ R01s) do so little teaching why do the public Universities insist that their Professors do so much research?
Perhaps Universities need to re-balance the roles of their faculty towards rewarding and encouraging teaching over doing research. Or at least equal.
Regarding the large number of comments on indirect costs, I would recommend a book for everyone here to read: “The Fall of the Faculty and the Rise of the All Administrative University”, by Benjamin Ginsberg, 2011. It puts into perspective the erosion of faculty and tenure systems, and their replacement by a behemoth administrative system of deans, deanlets, senior associate vice provosts etc. Deal with this bulky admin system, and research as an enterprise suddenly looks quite sustainable. Keep allowing this system to grow (at double digit levels) and we’re all doomed. Here’s the book…
http://www.amazon.com/dp/019978244X
I like the idea of capping the amount of funding received by one PI. Ones someone establishes his or her name, they get more and more funding to their laboratories while other meritorius projects from others lag behind. What is the purpose of having 4 RO1’s awarded to one PI in the similar area of research? The projects may be marginally different so that they can show no overlap.
In my opinion, the number one priority must be to limit indirect cost to institutions. The current system does nothing to encourage fiscal responsibility; to the contrary, when an administration wants to create a new position, they simply do so, and pass the costs on to the F&A calculation. Even the most rudimentary knowledge of finance will tell you that cash flow is everything, so there is, in fact, an incentive to have a bloated bureaucracy and bloated F&A rate. Some institutions (unnamed) have indirect costs as high as 99%, which means that for every dollar spent on research, a dollar of the NIH budget is not. I would recommend a percentage cap on each grant, as well as an absolute cap per institution;
In my opinion, limiting the size of awards is absurd. The cost of the proposed research is the cost of the proposed research; do you want to fund it or not? Where is that additional money supposed to come from if you limit the award? Stop fooling with the system. Peer review works. Let peer review rank order the applications, then fund what you can, starting from the top. On this latter issue, the very clear move towards bureaucratic and political decisions about what is or is not important is very dangerous. Scientists know what is important; let them decide.
There are at least two relatively ‘simple’ solutions:
1. Reduce the indirect cost for additional grants (50% for the second grant, 25% for the third, etc.). This reduction should limit the incentives for the administration to push PIs to write more grants.
2. Set a ‘minimum’ time effort that PIs can put on their grants. Realistically, to have a meaningful input into the project, you need to invest 20-25%. If you the NIH will use this number as a ‘minimum’, it will limit the total number of grants one can manage to 3-4.
ok, Dr G. And when you find that the big names / entrenched interests have 4 R01s at $469K in annual directs (or manage to get Program to sign off on one for $750K!) and the lesser names can’t manage to get past the modular barrier for theirs….is your problem solved?
It will not fix all problems.. but it will (a) automatically limit the number of grants per PI (3-4 is better than 32), and (b) limit the amount of indirects… Of course it will not solve other problems (big names, entrenched interests, etc.), but it might be an easy fix that could be implemented in very short time…
The discussion is great and one consensus is almost obvious, at least in this group: the current way of managing NIH resource is NOT working under current fiscal situation, and changes have to be made.
Let’s first make the message be heard by those people who can make decisions.
As to which options are the best, we can continue to debate. But to my opinion, any of them, or a combination of them, would be significantly better than the existing system.
The interests of scientific progress, medical progress, the research community, and the US taxpayer would be best served by following two basic principles:
(1) NIH support should be awarded based on scientific and medical merit.
(2) NIH support should be awarded equally to projects having equal merit and scope.
It follows from these two principles that:
(1) The NIH should award different direct costs to different projects (reflecting differences in merit and scope), but the NIH should not award different indirect-costs rates to different projects. The NIH should establish a single indirect-costs rate for all institutions, and this rate should be no more than 50-60% (and preferably should be less).
(2) The NIH should award different non-PI-salary costs to different projects (reflecting differences in merit and scope), but the NIH should not award different PI-salary-costs rates to different projects. The NIH should establish a single PI-salary-costs rate for all institutions, and this rate should be no more than 25% (and preferably should be less).
(3) The NIH should put extramural and intramural projects on an equal footing with respect to review and with respect to support. The current (extraordinarily) loose review standards and (extraordinarily) high support levels for intramural projects are unsupportable, unsustainable, and unacceptable.
(4) The NIH should not impose arbitrary caps on numbers or sizes of awards per PI.
There is an implicit assumption here that the proposal is the project. But the proposal is no more the project than the map is the territory.
We already have data showing that above $750k direct per investigator the ROI falls.
To make it increasingly difficult to obtain additional funding as this line is approached and then exceeded, is a logical and judicious approach that maximizes benefit to the scientific community and maximizes value to American taxpayers.
“The NIH should establish a single indirect-costs rate for all institutions, and this rate should be no more than 50-60% (and preferably should be less).”
As an FYI, such a policy would probably cause the Marine Biological Laboratory at Woods Hole to shutter, and Scripps as well. There are great research institutions that have no university affiliation to support the indirect costs with existing infrastructure. In my first comment above, I raised the question of whether the medical school research model needs to be discussed, yet here I imply the value of these soft-money-only institutions. But we think about these two situations very differently. Hearing the word “university” brings up a different set of expectations than “research institute”.
The combination of limiting the number of grants to 3/PI and capping the total dollar support to $1 million/PI/year will be the most effective way to support the current scientific community. Pumping money to a few big names will result in wasting money in the end though they could generate great science simply because they have a lot a lot more labors and resources than others, however, i don’t believe their output/$ is greater than those from smaller labs. In the long term, the low funding rate will hurt scientific community, let young generations lose their hope to do biomedical sciences.
Summary of main urgings (coded in line with slide 2 of the OER slideset):
A. The current way of managing things is the worst among options. “Bottoming out success rates” will not “allow the system to correct itself” in any constructive sense*.
B. Options for “change we can believe in” –
1a. Preserve and protect full funding of the direct costs of modular awards.
1b. Eliminate the principle of “hard” payline cut-offs for ALL investigators; don’t just “violate the principle of a hard payline cut-off” only for ESI / New PIs.
(i) preferentially fund the high-rated applications with lower total (direct + F&A) costs
(ii) prioritize highly-rated applications from those labs where the research has under-studied features (not just more chasing the same molecules or processes as are “hot”) and the total of all lab funding means that the lab viability requires the award.
(iii) Apply the cap on administrative costs in F& A at 26% of direct costs to ALL recipients of NIH funds if that is not the case already.
2. Place an NIH-wide “hard cap” on the number of RPGs and contracts on which one person can serve as PI. Three RPGs & contracts is about right (consider exempting R03).
3. substantially reduce the number of R01’s of sizes greater than 10 modules and larger RPGs (e.g., those over $500,000 direct per year); manage all applications for grants over 10 modules so that their success rate cannot exceed the pay line or the success rate of modular R01 applications.
4. (“5” – i.e, it’s not on NIH OER slide 2) Manage IC portfolio balance such that the NIH returns to spending ~39% of its funds on R01/R01-equivalent grants (exclusive of U01; U is not R-equivalent, it’s ‘cooperative’, i.e., managed), and 54-55% of its budget on RPGs.
* “Pax Romana”, ‘They have made a wasteland and called it a corrected system’ comes to mind.
AS an experienced reviewer who’s seen a lot of grants, I recommend:
1. Limit the total amount of NIH support that each P.I can assign to salary to 50%. this would make institutions pay for their faculty, who are often performing a multitude of other tasks and making the NIH pay for it. This would prevent institutions from over-leveraging their departments, which is part of the reason for growth/building boom.
2. Either cap total amount of NIH support per investigator (the tough love approach), or (a more merit based approach) at least utilize a sliding scale for success, where each new award has to compete at a lower percentile to be funded. (I think GM already does this.) I ‘ve seen enough grants from little labs, big labs, and even Nobel Prize winning labs to know that there is a diminishing return of creativity, brilliance, and productivity with each new award. (see Jeremy Berg’s wonderful quantitative study that corroborated this.) If we kill off a big fraction of the one RO1 labs to keep funding the multimillion dollar operations at their current levels, we will lose the generation of scientists that would have been trained in all of those smaller labs.
At the health sciences university where I am a faculty member, the “rich” disciplines (i.e., medicine) provide some salary support to their faculty for research. The “poor” disciplines, such as nursing, cannot do so because of the funding mechanisms built into health care. (Most nurse scientists cannot bill insurers for our clinical practice, and when we do, our reimbursement is less than that of physicians providing the same care. At health sciences universities, practice revenues typically fund research and education.) Thus, a disparity in research time exists.
In short, if I don’t fund myself, I get a full teaching and service load. If I can fund 15% of my FTE, I get a .85 teaching and service load.
Capping salary support for PIs on NIH grants would thus disadvantage the “poor” disciplines more than the “rich” ones. I am concerned that decreased NIH salary support would contribute to a “scientific silencing” of those who are already disadvantaged by the lack of alternate revenue streams.
First, I am delighted the NIH opened up this discussion.
I stronglyh agree with those who support a cap on indirect rates. There is simply no justifiable reason for the high indirect rates some insitutuions have managed to negotiate. I suggest that a 50% cap would be most appropriate.
I also agree with limiting the number of awards for any given PI. Absent a study that shows that the productivity of an individual with 6 RO1’s is 3 times that of individuals with 2 RO1s it would seem prudent to limit the total awards per PI.
Interestingly none in the debate have addressed the type of research being funded by NIH. In light of the current limitations on financial resources for NIH, and the prospect of further limitations, it would seem now is the time to examine carefully what areas of research the NIH should be funding. Each and every academic investigator and administrator can make strong arguments for their own particular area of research being a priority. But are blanket reductions across the research landscape the best way to handle shrinking resources? To insure that a priority is given toward research directly affecting the health of the American people will also require limitations on funding of research that does not.
The problem with this is that there are already too many people dictating what should be studied that have no business making those decisions IMHO. Congress isn’t willing to allocate more money to the NIH and doesn’t see the benefit in science, but wants to determine how much money should be spent on autism, or Alzheimers. (Not that these aren’t fine areas, of course.)
Basic science has produced important advances without being directed. The NIH funds research that is to be beneficial to public health and the scientists help decide which projects fit that bill and are good science. This system seems good to me.
In this regard alone I am actually a conservative and want less directional oversight. Fund good scientists to do good science and good things will come of it. Trying to make us all work on cancer or autism won’t help those with Parkinsons.
In the current budget situation, I think it is crucial to also modify the “two submissions and the project is dead” policy. A colleague of mine was notified yesterday that his A1 application with a 12% will not be funded (the A0 received a 15%). This is his only R01 and the current rules say that he can not even resubmit this as a new A0.
For the writers above who say modifications on how RO1 grants are paid will not be “merit based”. Can you honestly say that a grant scoring in the top 12% is not meritorious while a grant scoring 9% is??? That is certainly not my experience as a study section member.
If NIH still wants to keep the “only one revision” policy, I am OK with that. However, there still should be an opportunity to submit such a project as a new application. I also understand the concern about reducing the load on review so maybe sending in such a grant as a new application could be restricted to those getting 25% or better on the A1 review??? Either way, the current policy is not only failing to get excellent science funded, it is permanently shutting down projects found to be highly meritorious by peer review.
Maybe A1 grants scoring between the funding line to 40-50% have earned the right to an additional round of review?
The senior PIs don’t agree to lower the salary cap, don’t agree to limit grant number per PI, but hope NIH to award more R01s; the PI currently having a good score and waiting for FY12 approval don’t agree to decrease the size of award; young scientists having never been funded and eager for an independent research career hope NIH to give them more opportunities. All are understandable.
As a junior investigator, I had a couple of thoughts to contribute. First, I think that the debate regarding indirect costs is pretty obvious here. It needs to be a flat rate, and much lower than it currently is at many institutions.
Secondly, semi-selfishly, I am in concurrence regarding the limiting of awards per PI.
Effective management of funds, scientific “bang for the buck,” and effective mentoring of junior researchers seems to reach a point of diminishing returns after several awards. I can say this from my experience postdocing in a megalab where at least tens of thousands of dollars were wasted per year and several scientific careers ended due to a severe lack of mentorship. Again, there are exceptions to this rule, but I believe that they are the outliers in the grand scheme. I am not one to say what this limit should be, but it should definitely exist.
For those of you saying that this is an unreasonable request, I would say that instead of creating mega-labs, perhaps this PI funding limitation would create greater collaboration within our research community. When good science needs to be accomplished cost-effectively (I’m not saying without funds, just less), the good PI’s will find a way.
In the current budget situation, I think it is crucial to also modify the “two submissions and the project is dead” policy. A colleague of mine was notified yesterday that his A1 application with a 12% will not be funded (the A0 received a 15%). This is his only R01 and the current rules say that he can not even resubmit this as a new A0.
What makes you think that the A1 policy on revision was anything other than a response to the current budget situation? Did you miss the prior discussion on this blog after the Benezra/Noonan comments on the revision policy? Your colleague’s situation is indeed sad….but every indication is that this is the intentional outcome from the NIH perspective. They need to shrink the pool of applicant PIs. And this is not really a huge difference in approach. There were always near-miss scores in the grey area which Program passed over. Scores which were the PI’s only grant….or scores which would have become the noob PI’s first grant.
Why should we feel any more sorry for ourselves that this line is now in the 12-15%ile range instead of the 18-20%ile range?
If Program refuses to pick up a near-miss score from an established investigator who has a history with them….well, they are making a decision to reduce the pool. Or else they are deciding that this PI can survive until she comes up with a more interesting proposal. What they are not doing is making an error (according to their goals, not yours) because they had no idea what changing to a A1-and-out policy would do.
Right RockTalkBlogStaff? Surely you knew what all the implications were when you went down this road…
Agreed. The same thing happened to me a year ago, 14th percentile on an A1, and the project is dead. It’s one thing not to fund a project at 14th (or, in some cases, 9th…) percentile, but it’s just simply ridiculous to say that it is so inferior that it should be scrapped completely.
Perhaps you are right. The pool does have to shrink with the present budget climate. However, I would think that the pool is already shrinking so fast (and industry getting so many good people right now) with the experienced folks who have been well funded for years getting triaged or scores no where near the funding line that they could leave the door open to not shutting down labs which are making major contributions and whose science is judged excellent by peer review.
This hit home today since I just found out that yet another top notch scientist I know decided to take an industry position instead of staying in the soft money rat race. I know this is going to just get more common, but I hope that some mechanism can be found to preserve the best academic science (not just the projects lucky enough to score in the top 5-10%). I guess the up side is that our biotech industry will be getting some really smart people in the near future.
I’ve been well funded for about 15 years with previous scores well below 5% ;however, it looks like I am now one of these “old” investigators NIH is aiming to cut. Too bad because I have spent a lot of NIH money to develop some complex in vivo model systems that are now to the point where I can get a lot of mechanistic data. How do I get one of these Industry jobs? I am close to 50.
This is an excellent point about the typical career arc. My completely subjective view suggests that there have been, in the past, many good investigators who labored mightily for their first 5-10 years trying to build funding, staff, capacity, models and a scientific program. When they hit their mid40s or so, things reached the exponential phase that last about 15 years. The models were humming, the research projects were turning into programs, the trainees come a-running and, in the past, they were able to load up on a good deal of sustained NIH funding. A lot of great papers resulted. In part because they no longer had to really worry about getting the next grant when they needed it and in part because they were able to compete for multiple awards and a fairly hefty amount of total support.
Now this demographic is squeezed. The proposals in this thread to soak the rich / cap total awards prevents the mid-career scientists of approximately GenerationX from blossoming like the previous generation was able to do. Yes, even though they are meritorious because of the aforementioned development of their programs. The emphasis on the ESI (which is a good thing) further saps their ability to fully realize their efforts. And the top end of nearly emeritus faculty is refusing to retire, slow down, scale back their laboratories, etc, all the while demanding Bridge funding from their friendly POs when things go wrong with study sections.
I ask you, is the research enterprise well served by throttling back the very demographic that is in what has, historically, been the most productive interval in a scientist’s life?
Don’t kid yourself. Big cuts are coming to industry as well. As the government looks to cut the deficit by reigning in health care costs, the big pharma and the biotech industry will be hard hit.
I definitely support capping indirects. It’s absurd for some institutions to receive indirects at 150% or more. Cap at current median indirects – likely 50-ish%. Big Name and Politically-Connected insitutions protest that they lose on this? Boohoo – they can hit up their alumni for more endowed chairs. I also agree to capping the number of RO1’s per PI – I’d say at two – as the ‘mega-holders’ can compensate with PO1’s, etc. But having reviewed applications from megaholders, there is always major undeclared overlap that Program lets slip by.
Don’t limit the awards to any individual or institution, research grants are not am entitlement but should be merit and science-based first and foremost. But-there should be a mechanism for funding projects, esp CER, that are, overall, net positive due to possible savings in healthcare costs from the Medicare program
A (very) modest proposal: stop funding investigators at *foreign institutions* with R01s. It won’t solve any budget issues but it will address the bizarrely inappropriate largesse of funding foreign – mostly Canadian – investigators at a time when paylines are below 10%. Why are there 27 R01s awarded to investigators at the U of British Columbia right now? Wholly irrational, and at the most basic level, it’s 100+ science jobs that we’ve shipped over the border to Vancouver.
In case you’re wondering, it’s not a mistake the Canadians are making in their funding decisions. Here is the eligibility criterion for NSERC funding:
University faculty: To be eligible as an applicant or co-applicant, you must hold or have a firm offer of an academic appointment at an eligible Canadian university at the time of application.
An excellent point that I have been raising for some time now. I don’t know of any other major foreign granting agency that provides taxpayer funds to foreign institutions. I received my first R01 a couple years ago and was given a list of the other first time grantees. Among an already short list of awardees, I counted several R01s going to institutions outside North America. Particularly in the current fiscal climate, it seems like a no-brainer to FIRST eliminate foreign awards. Next, we can have a discussion about what additional cuts need to be made.
What happened to basing the funding on merit? Nationalism does not make better science.
One must ask the same question to the funding agencies from every other country. I have reviewed grants for foreign granting agencies. The basic stipulation for eligibility is to conduct research in that country. Likewise, all of my colleagues who are PI’s in the US, but citizens of another are no longer eligible for funding from their home countries. With such a situation, I find it difficult to justify letting such a large amount of meritorious research in conducted in the US go unfunded, while at the same time sending research funds to countries that as a rule do not fund foreign research. Certainly, there is no shortage of innovative research ideas within the US that are worthy of research funds. Isn’t that one of the major reasons we are having this discussion in the first place?
I agree 100%. The US makes it very easy for researchers to come to this country. If you want to do NIH financed research, you can do it here…
It might not be a substantial amount but at least something to fix without affecting researcher in the US.
Why do the proposed options to manage NIH resources exclude the costs of grant administration at NIH? Do we really need that many research officers etc. at such high salaries to essentially push grant papers around? How about cutting the number of these positions significantly and swiftly re-direct the saved money into research funding?
Cap the indirect costs because they are the legal way of institutions to milk (and rip off) the government. Every foundation caps indirect costs, why not NIH? 50% of NIH funds go to only 20% of the investigators? Wow, even these “star scientists” cannot honestly predict that their research will bring the next break-through in science because discovery research is a high-risk enterprise. Therefore, NIH would be smart to spread the risk of its investment among MORE RO1 investigators.
This could be achieved by demanding a significant minimum amount of effort (e.g. 30%) a PI must spend on a project.
Limiting the amount of total funds per PI is another good option and would reduce the support of “Manhattan projects” by NIH. For example, clinical trials are too expensive to be supported by NIH. What has become too risky and expensive for many Pharmaceuticals, should not be funded with taxpayer’s money, certainly not in these dire times.
Capping the salaries of PI would mainly affect the physicians in research because very few PhD scientists will ever get close to the existing salary cap. I am in favor of this cap because, apart from saving money for NIH, it would remove the current salary discrimination against PhD scientists by NIH.
Well, Lorenz, if all POs did was push grant paper around, as you assert, then I should be replaced by a GS 10 clerk tomorrow. The fact is that my colleagues and I spend very little time pushing any paper around at all. Most of our time is spent advancing public health. Administratively, we help PIs respond to unanticipated problems and discuss new projects to move the field forward. Scientifically, we spend time recruiting talented investigators (both new and seasoned) to apply for NIH grants – often bringing in talent from outside healthcare to solve public health problems. We also write journal articles to help the research community learn about the nation’s public health research needs, and promote awareness of NIH research findings that might inform ongoing studies or encourage new applications for research grants. As public health officials, POs also spend a lot of time helping get research implemented into everyday healthcare by meeting with outside government agencies and helping bring together private sector healthcare providers and grantees. POs also keep IC management informed so they can keep Congress informed. Now and then, we do progress reports and other administrative stuff, but that work contributes to accountability — not so much to improving public health. The diversity of support POs provided NIH not only requires a doctoral education, in most branches is also requires decades of research and implementation experience before moving from the bench to grants administration. Are we overpaid? Not based on the job offers I get from outside NIH.
Here are few suggestions.
1. The institutions should invest at least 50% of the salary. This way institution are responsible for the success of the PI. In the absence of institution commitment, these PI become a parasite on the NIH system and submit many grants to support their salary and basic necessities and overburdens NIH system.
2. Administrators are generally not interested in the PI’s science as much as they are interested in indirects. If the indirect’s are reduced by few dollars everyone is after the PI. If you need to get the science done, you need to reduce the indirect costs may be as low as 30%. The existence of this blog is manifesto that there are many willing and meritorious PI’s who are willing to take funding and do science. So the argument that institution would not conduct research if indirect’s are reduced is not a valid argument.
3. Most of the time a PI feels that he/she is working for the indirects and not the direct cost of the budget.
4. The PI salary should not be charged multiple times from the tax payer to facilitate the other functions of the enterprise. Specifically true for state institutions when salary is a line item in the state budget and is also charged to the grants. Leaves a lot of salary, fringe, indirects for administration to fund other operations as that money is never returned to the state.
The government has invested a considerable amount of funds in training a multitude of scientists that cannot maintain their labs through these difficult times without an adjustment of the status quo. To prevent the loss of this investment, our goal should be to keep productive, creative labs in business until the crisis is past. This may mean limiting the number of grants a PI can have, limiting NIH support of salaries, and limiting indirect costs. This may even have the effect of increasing the return on the goverment’s investment. Limiting the number of grants a PI can have is not necessarily equivalent to stifling the most creative and productive labs. We all know that there are many less objective factors at play in grant review, and productivity seems to plateau as funding dollars skyrocket.
I’m late to the discussion, but looking at the slides and having thought about this a lot I think the best solutions are:
1. Ask medical schools/institutions to kick in at least 50% salary. They can take some of that from the indirects.
2. Limit the amount of funding for a PI to $800,000 and/or limit the number of grants.
3. The indirects are outrageous. I get that the University needs money to do maintenance and keep the lights on, but why should the scientists shoulder that load for the rest of the departments that don’t bring in grant money? (At least at primarily undergraduate institutions).
I see people gripe about this not being “a merit based system” if we implement restrictions on funds, but I disagree with that idea. My science is solid and I publish well and yet I am struggling to get enough money to run my lab. 1 R01 is all I need to keep going, but it will never be enough for me to hire 20 postdocs to run a factory. It’s barely enough to hire one. So yes someone with more than $1,000,000 a year can outproduce me and my 200K budget because they can pay for next gen sequencing, and multiple postdocs, and microarray experiments. They will continue to look more and more stellar while I appear to be solid but not adventuresome. And this comes back into play because I can’t compete with that on “merits”.
That’s life, but I resent the idea bandied around in these comments by some that somehow because I am struggling, it’s all my fault because I must be a marginal scientist. I don’t think we should give money to bad science just to be fair, but spreading the money to the scientists the NIH trained in our graduate programs a little more inclusively seems like a good idea for everyone.
There are clearly many opinions about indirect rates. How about incentivize institutions to work to reduce their indirect rates? For a start non RO1 grants could cap indirect rates, and any administrative supplements and other handout style awards could set preference thresholds so that fiscally well managed institutions with lower rates would have a better chance for those supplements. The goal would be to wean institutions from extremely high rates and eventually over several years drive rates down.
The other thing that seems insane is the end of the year binge spending that goes on in grants. Anytime significant funds are rapidly spent in the final month of a grant this is a waving red flag for mismanagement or at least that fluff and excess was built into a budget. I hate seeing large sums of NIH money being rapidly and sometimes wastefully spent. This needs to be put under control…good luck
There should be 1 RO1/PI cap so that everyone gets a good chance of getting at least one RO1. No PIs should be allowed to apply for another until 4th year of the term.
I am against reducing grant sizes or capping money per researcher. What has made our research enterprise successful is an emphasis on merit. When money gets tight we should protect productive researchers and discourage the relentless expansion of PI-level researchers. As far as the current mess that represents funding, the R01 system was designed for a time when one was sufficient or, more recently, when getting multiple R01s was not time-prohibitive. Long-term I support an HHMI- or Welcome-style system where PIs are judged every five years based on strictly on productivity. Want to try something risky? You have five years to get it to work and then you’re judged. Want a larger lab? Be more productive and apply to move up to a higher ‘track’. This will allow the system to plan ahead, determining how many PIs it can support across time.
I dare say you need not worry about a lack of discouragement to the “relentless expansion of PI-level researchers”. How many non-foreign born graduate students or post docs do you have? How many express an interest in academic science? I recently posted a job for a technician. I had 3 newly minted Ph.D.s apply. When I asked why-they pointed to the lack of jobs, the potential for a career as a Post Doc, and the non-viability of an academic science career given the current problems with funding. The current system is unsustainable and needs to be changed when we suddenly no longer fund projects that are judged outstanding and excellent. You’re not giving the chance to many excellent and outstanding PIs to become those “productive investigators”.
“How many non-foreign born graduate students or post docs do you have?”
7 of 7, at the moment.
I believe that it is a great idea to cap the number of grants per investigator or set up a requirment for minimal effort (e.g. 25%) per grant. I participated in many review meetings and realized that the productuvity is not proportional to the number of grants or the size of each grant. Ther are too much waste in big labs while others are struggling to keep their labs. The science as a nation cannot just depend on few labs. Creativity comes from cometition in big pools. I also agree with the idea that indirect cost should be reduced, particularly for those institutions with more than 60% indirect cost. There are excessive administration staff and salary in many institutions which can be significantly reduced. For instance, the salary of our adminidtrators got doubled or even tripled in the last couple of years. We have no idea how indirect cost was distributed. These issues should be discussed at various levels in NIH.
I’ve read a number of comments here, but none seem to be addressing what I see as the main issue. How many scientists do we need, or want, as a society? The current system is unsustainable – the growth in the potential number of PIs is exponential and the dollars are not. To some extent, we can force ourselves to do more with less. But that only goes so far, and reduces what we can do. The main message is clear, either the funding stream needs to increase, or the number of scientists needs to decrease. There is no point in training graduate students and postdocs in either a big lab or a small lab if they are not going to find a job. The mission of NIH is to get the science done. This is not the same as NSF, which has a training/education mission. NIH should continue to fund the best science.
I hate the idea of losing excellent scientists. But the system simply cannot support the number of scientists we are training. While the recent programs that benefit younger PIs are great (for them, at least temporarily), they are nothing more than Band-aids, that push the problem off a few years. As was mentioned by others, what happens at renewal time? No favoritism then. Yes, most scientists do their most productive, creative work in their younger years. This doesn’t mean that they are useless to the research enterprise as they age. Or are they? Is this the goal? Get a good ten years out of them and then hope they fade away? My gut tells me the young PI initiatives were put in place with the expectation that things would get better by the time renewals came around. Well guess what…? In fact, the young PI initiatives are making things worse. Numerous young faculty get jobs and tenure, and then are firmly in the system…and then research funds dry up. We need a sustainable model. A good place to start would be to eliminate the ideal of infinite growth. In other words, REDUCE THE NUMBER OF NEW FACULTY HIRES! Yes, hire enough to sustain the system and the scientific enterprise, but that’s it. Supply far exceeds the demand. NSF and NIH should work together on this to reduce the graduate student pool. So, who will do the work in the lab? Well, I find that the best most productive scientists are senior postdocs, or super-postdocs/mid-career scientists. Instead of sending them to industry, why not keep them in academic labs? Yes, they may cost a bit more, but they are much more productive.
Now as depressing as the above may sound, I hope that we don’t go that route. I’ve heard the arguments that we need many more STEM graduates. But the only way society is going to get these people is if there are funds to pay for them. This gets back to what I said at the beginning – how many scientists do we need, or want, as a society? Otherwise, it’s just a bubble, and it’s bursting now…Write your Congressmen/Senators and ask them what they see as the future of the medical establishment.
I think this is the philosophy behind the 2 strikes and you are out policy. The research community has overwhelmed the system, both for funding and review. It’s not sustainable without continued increases in NIH funding. I actually do think the pool needs to be trimmed and the number of researchers decreased. The most interesting comments on this forum to me have been those recognizing the difference between NSF and NIH. We don’t see the proliferation of soft-money positions in the non-biomedical sciences. There are no expectations on the part of NSF grantees that they can completely support themselves and a small army of technicians, students, and research associates on grants. The research enterprise in the biomedical sciences have become completely distorted into a money-chase.
The current system appears unsustainable, so we must cut back. Our goal should be to get the best science/dollar ratio. So,
(1) The peer review system should take the cost into account in rating a proposal. The “cost” should be the total cost, including indirect rates.
(2) As best as we can manage it, the “merit” should be real, not based upon the reputation of the scientist or institution.
(3) If proposals at, for example, the 8 percentile are indeed nearly indistinguishable from those at 12 percentile then let us treat them as so. Let us give all worthy proposal the same opportunity for funding. We then break the tie, by a throw of the dice, by apportionment according to the number of senators plus house members for each state (as is common in Federal funding), by a goal of diversifying science, by a goal of focusing on the most important areas, or by some combination of the above.
If you take away all the ways that the NIH balances the scales, your “merit” approach will result in further concentration of grants in a few coastal elite Universities. In the hands of molecular biology / -omics mega labs chasing basic monstroscience in an ever demanding race to beat the other few guys to that Science or Nature paper. Yep, even on a bang-for-buck basis. (like the mega labs couldn’t adapt and win every possible “productivity” game you choose to enact?)
Is this what you intend? If not, your calls for “only on the unbiased merit” are not serious.
Sometimes grant proposals that are out of the mainstream do not fare well in study section due to lack of reviewer expertise. With paylines at 10-12% (instead of 20%), these grants, which can represent very new and important directions, often have a difficult time landing within the funding range. I agree with a previous comment that when 25% of the funds are allotted to 5% of PIs, this limits creativity and diversity.
I have following suggestions regarding NIH funding:
1) NIH should prioritize RO1 and other individual grants in the current funding situation.
2) To effectively use tight budget, NIH can require minimum effort for PIs in each grant. For example, there is a 50% effort requiirment in the Director award.
3) To eliminate and reduce large program projects. These projects are costly and less productive based on the recent data. There are too many “big” scientific administrators who hold significant amounts of funding but barely participate real experimental designs and data analyses due to their busy schedule.
4) The funding decision should be always merit based rather than be handled by POs in ICs. The sudden funding policy changes made in NCI this year has negatively impacted many dedicated investigators and generated unnecessary anxiety and confusion in cancer research community.
Why do you assume that PO priorities are not “merit” based and why do you assume study section ranking is a purely “merit” based ranking?
In times of tight resources emphasis should be on maintaining as much of our nations science infrastructure and manpower as possible. This will require restricting the number of grants per investigator to two and capping salary support at 50%. It would also be useful to fix overhead cost rates to approximately 50%. Some special funding initiatives should be available as well for especially innovative large programs but in general funding for program projects and contracts should be restricted.
So it is better to preserve the infrastructure (For when? What future?) than to get the science done?
What makes you think the science wouldn’t get done under Richard Roman’s plan?
Wasn’t it you yourself that pointed out that certain types of science (which for better or for worse are at present times thought of as the hottest and the best) required an army of postdocs and lots of cash to spend on -omics core charges? Someone else above pointed out that smaller research Institutes, including some of the very top producers of science* that we’ve seen in recent decades, would be dead as a doornail under the Indirect capping proposal.
It is disingenuous to pretend that major changes in policy are not going to result in major changes in the scientific product of NIH’s extramural efforts.
Far better, and more honest, to recognize the impacts, discuss them openly and then to choose alternatives in full knowledge of the best prediction of likely effects.
*Scripps just scored a Nobel Prize, no?
My question was really a need for clarification. I *think* you and Richard Roman are talking about completely different things? You said you are against preserving infrastructure, but then argue against capping indirects?
My post above was a reaction to the posts where people state they are worried that any proposals we put forth to change how money is budgeted would eliminate “merit” as a criteria for funding. I think it is already hard to measure “merit” between a lab with little to no resources and one with more than they need.
I don’t think I said I’m against “preserving” infrastructure? I *do* believe there are trade offs- and the soak-the-rich comments, including yours, seem to ignore the fact that mega-lab science is that which we currently view as the best science. It respects “merit” to continue to award certain labs an armload of R01s.
This is where I object to comments like RR’s. They assert that mega-lab science is worth dismantling so that more one-grant labs can stay alive. Nice from a labour perspective but it is hard to see where this is obviously better for scientific discovery.
You and I possibly agree on the point where there remain some inequities in review in which approximately equal Old and Young PI proposals get different outcomes. I would support attempts to level the field at the study section level.
1) All this talk about F&A is irrelevant because these rates are negotiated with the institutions’ cognizant agency (DHHS or ONR), and OMB has made it clear that agencies can not demand cost share, which is what they would be doing if they tried negotiating F&A rates on grants. (There are exceptions, including training capped at 8% and USDA’s weird calculations- all congressional decisions). I think NIH is looking for suggestions on things they can actually control.
2) Choice of funding mechanisms (R01, P01, etc.), capping the number of grant awards to a PI (or an organization?), and considering the use of targeted (RFA) versus solicited are things under NIH control, but how about thinking creatively about changing research paradigms? Should NIH move more upstream or downstream in their funding focus? Perhaps there should be more emphasis in NSF funding basic research and NIH on more applied (clinical and population). Reduce training and career grants and focus on research? A major change would be to switch to a cost-reimbursement model rather than a prospective project model. (probably something only Congress could decide.) Perhaps grant applications should include a plan for outreach to unfunded researchers to support networks of collaborators to spread the money around. Can some of these massive lab-factories become regional research resources for smaller institutions?
3) I replied to a comment above that I thought the research community is too large to be sustainable, even with continued increases in NIH funding. We have seen the problems with allowing financial institutions to become “too big to fail.” Now are we propping up research institutions of the same ilk?
Agree that forced renegotiation of F&A seems unlikely, but NIH CAN control which grants they select for funding, and do so on the basis of whether the science of an application with way higher F&A costs is so unique that it would justify selecting that one as compared to another. Market forces will do the rest after implementation of such a selection factor. [As it stands, taxpayers are being asked, in effect, to pay for Lexuses in the F&A world when the study sections have said an Accord would do just fine. Indeed, it would take more granularity about the nature of the awards, but the data showing that success rates improve as the size of the award gets bigger suggested that there is at least a bit of low-hanging fruit that would be politically wise for NIH to trim.]
In order to bring much-needed fairness and transparency to the NIH funding system, I suggest the following:
(1) NIH should cap the dollar support to each PI at $400K/year. Additional money is just a waste as it rarely leads to any improvement in scientific productivity.
(2) NIH must abolish all indirect costs and leave the responsibility of providing research infrastructure and administration to the awardee institutions. The indirect costs simply allow these institutions to put their money (that could be used toward research infrastructure and administration) either into the pockets of deans (the one percenters) or simply waste them on paying salaries of unnecessary administrators and bureaucrats, whose job is nothing but interference with the smooth flow of good science.
(3) The awardee institutions must pay 100% salaries of PIs.
(4) An institution that does not meet the above requirements should be considered unfit to conduct NIH-funded research and should not be funded with NIH money.
Just some thoughts! But I know it will take a lot more than just some thoughts to make NIH a fair system. In the meantime, let the devil rule. The power lies with corporations that the medical schools and universities have become not scientists. These corporations are the masters and we the true eternal slaves. Either follow the orders of your master or simply perish.
As noted above, this–“NIH must abolish all indirect costs and leave the responsibility of providing research infrastructure and administration to the awardee institutions.”– means that Cold Spring Harbor, Scripps, the MBL at Woods Hole and many, many other research institutes would have to shut their doors. And that’s just by abolishing indirects. Your other suggestions have similar unintended consequences.
(Also, fewer than 10% of university deans are in the top 1% salary range. Most of them make well under $200K, and fewer than 10% are anywhere near $250K.)
I like the idea of a $ cap/investigator, that way the PI can prioritize their goals.
I don’t agree with comments on overhead. Someone has to pay to light and heat my lab. Someone has to pay for staff managing all of the federal regulatory and compliance issues.
I also don’t agree with the institution paying faculty salaries. Why should they do that, what benefit to they derive for paying me to work on research mission instead of teaching and service missions. But I do think there should be a cap on PI salary contributions with the institution contributing perhaps at least 50%.
It’s discouraging to hear comments saying that we SHOULD limit the number of people coming into science. Many in our society (and probably all of us in this discussion) think that science and technology are critical to maintain a competitive edge in the realms of economics, politics, and innovation. As such, broad efforts to attract students into STEM disciplines have been ongoing and largely successful in recent years. However, the current funding NIH system (though I think this applies to NSF and other agencies as well) sets up a pyramidal scheme, where far more are recruited into science than have a hope of surviving. As a (so far) successful PI, I have been asked repeatedly by graduate students at my institution and those I visit “What kind of careers should we look for since there aren’t jobs as PIs?”. There is a widespread and becoming-true sense that there isn’t much of a future in investigative science. Is this really what we want?
Darwinian evolution works beautifully for actual evolution, and “many are called, but few are chosen” is a fine mantra for many fields. Science that is funded by taxpayer dollars should absolutely be competitive, to ensure that our money is well spent on research that is deemed significant and appropriately conducted. However, in the current climate, everyone acknowledges that truly excellent science simply isn’t being funded due to budget limitations; many here have commented how a funded 6% grant is indistinguishable from an unfunded 14% grant. I hear many colleagues on study sections lament that the top 20% or so of applications are all terrific, but with paylines of ~10% or less, most go unfunded.
This is a serious long-term problem, and NIH resource management strategies could help. The current approach will ensure that the minority of investigators who receive most of the money will thrive, while we not only lose the rest but discourage young people from entering our field. Have we gone to so much effort to recruit youth into science only to now close the door? It will be phenomenally difficult to bring them back again when our budgets improve down the road.
Imposing some kind of limits on funding to individual investigators seems to be the most viable survival strategy. One good example in the OER presentation is limiting PIs to two R01s, which would add nearly 1000 funded applications. That represents nearly a thousand PIs who would stay in science, or new PIs coming into our field. NOte that at current paylines, this doesn’t require funding “anyone who wants to do science”; there are plenty of excellent applications just beyond the paylines waiting to be funded. Similarly, shifting the allocation of NIH funds from large contracts and centers (which largely give more money to those who already have it rather than funding individual scientists) to individual or multi-PI RPGs would also support a broad foundations of researchers. Individual basic and clinical researchers are the life blood of investigative science, and keeping more of them in the field is a better strategy for surviving and thriving than completely letting the current “market” forces persist.
Two simple suggestions that would create incentives for the Institutions to reduce indirect cost:
1. Among grants with similar merit scores, the grants with lower indirect costs get funded.
2. Indirect costs for multiple grants originating from the same PI are calculated as follows: 100% -the first grant, 50% -the second grant, 25% the third grant etc.
As a reviewer who has served on many CSR study sections over the past several years, I would like to emphasize that the NIH grant review system is completely broke in that it not only favors the “Winner Takes All” approach but also fails to reward talent on a larger scale. Most reviewers would agree that in any given study section, about 10% of applications stand out from the rest for their extraordinary merit, while another 10% could be easily triaged. But the remaining 80% of applications are virtually indistinguishable in terms of their scientific merit and the task of differentiating them for the purposes of funding is extremely subjective and constitutes nothing more than picking a winning number from a lottery draw. In these times of limited research dollars, it is more incumbent than ever that the NIH grant review mechanism stops acting as a national lottery draw and finds a more objective metric for funding scientific research.
I would like to suggest that the balance between intramural and extramural funding be re-evaluated. There seems to be little control of the intramural support. As an example, some NIH investigators keep huge mouse colonies that they are allowed to grow by breeding without the PI knowing what the mice are, let alone having streamlined plans for their use in an economically justifiable scheme. Many extramural scientist on the other hand have to eliminate valuble mouse strains because of lack of funding.
I support enforcing a lower salary cap for all investigators, possibly based on institution location, as well as enforcing a salary cap on research associates, technicians and students covered by the grants. Often, the PI is not responsibly using the NIH funds when it comes to salary levels for the members of his research team; this is driven in part by competition and the fear of losing a research team member; a salary cap for everybody could solve this problem and significantly reduce the overall costs. Also, we should stop allowing direct costs for tuition; at my institution tuition is covered by indirect costs.
Maybe productivity should be analyzed relative to costs. Small research groups may produce much more than large groups when productivity is analyzed after normalization by the amount of invested dollars.
Last, I fully support revisiting in depth the management of indirect costs.
When the NIH budget was doubled, there were two trends that now haunt us: 1) the large research Universities built more buildings, all of which have to be maintained, and filled with researchers. 2) They hired soft money researchers, expecting them to be self funded in perpetuity. Both mean that responsibilities for maintenance and for faculty salary got shifted from the Universities to the NIH. In most countries in the world, Professors get a salary that is stable, and a few technical staff and/or students/postdocs to be able to do a limited amount of work without extra funding.
It is time that the Universities are told: take responsibility! If you hire someone, he is yours to mentor, foster and pay, and your risk. So, I applaud Francis Collins et al. who are saying 50% of a Professors salary has to come from the University – or a maximum of 50% can come from NIH – whichever. Just like the economy and the housing boom, the Universities relied on an ever increasing NIH funding availability, and that is wrong. Although I now have an “outstanding” grant that will likely not get funded, I don’t sign the many calls for asking senators and representatives to not cut NIH. I appeal to all the professional organisations to appeal to their own Universities to take back the responsibility for their faculty, to end the spiral of unsustainable growth.
It is not essential to have 80% indirect costs. So, a nonnegotiated, flat indirect cost rate of maybe 35% seems easy to implement – there is no right a University has for NIH Dollars, so they can’t sue NIH. They can move some indirects into directs, but fair enough. We know that most Universities give kick-backs based on indirects, so they aren’t ALL necessary!
Another suggestion is a different measure of productivity: Papers per $100K. As someone above said, rich labs use a lot of money but not judiciously, and getting really good science done at low cost is better than very expensive science. That said, of course, clinical trials, and certain animal experiments are inheritantly more expensive. But some papers are also harder to get – so the correct metric of papers/grant, when people with multiple grants are allowed to double dip (mention several grants in each paper), is not fair either.
I am completely against your last metric – papers per $100K. The fact is that some types of work are cheaper than others. That does not make them better. And a Cell paper is most likely better than one in many of the obscure journals popping up all over the place.
I suspect the reality is that for universities (& research institutes) to have to cover & guarantee, say, 50% of salaries & fringes from non-NIH money (sources other than sponsored research), the less positions there will be. Not sure if that’s what everyone on the forum wants, or what’s best for the nation. Similarly, the more institutions with limited resources are forced to commit, say, 50% of salaries and fringes, the more pressure there will be to move the bar on tenure very high up.
I believe that number of papers per grant is not good enough ‘metrics’ for merit grade, or funding, though that number do feed into the publication industry, or maintain it.
it is more incumbent than ever that the NIH grant review mechanism stops acting as a national lottery draw and finds a more objective metric for funding scientific research.
Your analysis is correct but your plea for improvement fails. Because you are confusing an inability of the system as it is now to distinguish with the reality that a very large percentage of submitted applications really do not differ on any possible objective metric that we might think of as scientific merit. As assessed in advance of the conduct of the work, of course.
There is no objectively better way to select the subset of proposals the NIH can afford. Pursuit of your pipe dream would be a dangerous distraction from the real job at hand which is trying to set priorities. Priorities for who is going to take the hit and who is going to benefit. Which way will the study section table slant? Will we try to keep the current infrastructure alive? or try to shrink it down by a significant amount? New Assistant Professors? or shall we eke out another 5 years from those nearing retirement? Basic, translational or clinical? Are we after therapies or Nobel prizes? Incremental advance? or big-splash papers that might turn out to be completely unimportant later?
IMO, this is what we should be discussing. Priorities and hard choices. Not pipe dreams of some huge pot of gold hoarded by the Indirect Cost Goblins…
Exactly.
Program administrators have been maligned on this forum, but most of the ones I have met were very smart people. (Admittedly, some were arrogant jerks.) Their job is difficult and will get harder. Are all the applications scoring in the 5th percentile really that needed? Or are they proposing the umpteenth mouse model that looks at a molecule in a well-described pathway? Perhaps the review criteria need to emphasize novelty. Funding in rank order may not be a good strategy anymore, so we may see the published paylines disappear. We may be seeing published lists of areas that institutes will view as low priorities. The research community should be vocal on setting those priorities or they will be set by NIH.
To address potential solutions to the current and evolving funding situations, the following approaches may be worth considering:
1. Emphasize more on partnering between pharmaceutical and biotech companies with academic institutions and NIH to fund basic and clinical research.
2. Establish a process for co-funding between the NIH and industry.
3. Home institutions should consider paying a major portion of salaries of established investigators, who have shown an excellent record of scientific contribution and continuous external funding. This will free up federal grant money (salaries) for more research and/or supporting more grants.
4. Consolidate various programs at the NIH to increase RO1-type grants and monitor research outcome of R21 and RO3 and other big programs, and if necessary, eliminate these categories.
5. There needs to be a more efficient use of NIH funds. For example, indirect costs may be on a sliding scale for individuals with multiple grants and re-direct those funds to pay new research grants. It is time to seriously consider limiting the number of major grants an investigator can hold. Also, productivity of individuals holding multiple grants should be critically assessed in terms of scientific contribution against dollar amount.
6. It is time to eliminate bureaucracy and multiple layers of administrative responsibilities within the NIH and home institutions for administering grant review process and funding to reduce administrative costs.
7. Build strong grass-root advocacy groups for public and congressional awareness to the importance of basic and translational research and build a strong campaign for increased NIH funding.
8. Start with a “Million Scientists March” to Washington, DC, involving scientists, educators, graduate students, postdoctoral fellows, research personnel, people from biotech and pharmaceutical companies, and citizens who care for scientific discoveries that improve human health and quality of life. This approach will educate the society at large, draw the attention of decision-making bodies and raise the stature of scientific research and the benefits society reaps from it.
“It is time to eliminate bureaucracy and multiple layers of administrative responsibilities within the NIH and home institutions for administering grant review process and funding to reduce administrative costs.”
The problem with suggestions like this is that the bureaucracy grew in response to 1) the need to comply with federal regulations around the use of federal dollars, and 2) the need to complay with calls from Congress for oversight of taxpayer dollars. Any of you who have ARRA funds and complain about the quarterly reporting requirement should think about the fact that those reports represent the tip of the iceberg for the kind of accountability that is mandated, and undertaken by those bureaucrats, for federal grant dollars.
Completely true about the auditing infrastructure growing because of fears that someone would be “stealing” or inappropriately using grant funds. Yes, some of the high profile cases involving federal grant funds that have come to light in the past decade are shocking, however, I don’t see how they could ever happen at an institution with minimal accounting controls. Certainly my institution is on top of this to an extreme extent.
I came to the realization that the auditing has gone out of hand when I was site visited (a day visit in my lab) by an NIH official because I received a $4,000 ARRA supplement to support an undergraduate. They interviewed the student extensively and my entire lab about his project. I was greatly puzzled by this but the program officer said that they were required to site visit these awards to ensure I did not just hire my nephew or something. Really, for a $4,000 (not a misprint) award??? How many more awards could have been made without the site visits?
Even more recently, as part of an audit of one of my NIH grants I was questioned about a $20 charge for five bottles of nail polish and had to write an extensive essay justifying why we use this in my research (and yes we really do). (and no, I am not being targeted due to some prior impropriety, I have had completely clean accounts for 15 years)
Yes, we don’t want NIH funds misappropriated, but here the auditing costs much more than even any potential misappropriation. Something has happened to any possible sense of perspective here.
Seasoned Reviewer wrote that he/she has always wondered how institutions whose faculty is entirely on soft money can afford to put up so many new research buildings. In fact, institutions can afford new buildings precisely because of the money they save by not paying salaries. This starts a positive feedback cycle that siphons big money out of NIH because when new buildings and other facilities are added, higher indirect cost rates can be negotiated with NIH. NIH may want to reconsider this policy and put some breaks on this “death” loop.
I am not in favor of limiting the number of awards per PI because some grants are small, like RO3s, which only provide $50K/year. I confirmed with NIH Resource Management that their grant limit calculations included small grants.
Limiting the amount of $ per year makes more sense. When labs get too big, the return in research quality can diminish. I checked with NIH Resource Management to see whether the numbers they ran were based on direct costs only or direct + indirect and found that they include indirect costs. So a million bucks a year for a PI boils down to about 2 R01s + an R21, assuming indirect costs of ~60%. That sounds pretty reasonable. I’ve always been a fan of quality rather than quantity in science. Limiting $ forces PIs to focus on their best stuff and levels the playing field.
How does a hard cap on total $$ level the playing field when different types of research cost widely different amounts? People doing chemistry would be awash in largesse while those doing human clinical trials wouldn’t have enough to do much. Fly and worm genetics would squeeze out mouse genetics. Etc. Cost/productivity analysis sounds great, until you admit that it is just one more way to reshape the *kind* of science that will be done.
If that’s your intent, fine. But be honest about it.
@ Old Timer — Thank you for seeking & posting clarification of a few key points.
Definitely, while it is vital that NIH implement limitations of various sorts (even if the “Gang of 12” / JCDR does reach an agreement and NIH is spared the 7.8% automatic cut or a cut from the JCDR agreement), the changes have to be done carefully and thoughtfully. Small awards should be excluded from a cap. Unfortunate but understandable that they used total (direct + F&A), in which case the ~$400,000 cost is a 10-module R01 with ~ average F&A. It should not be just two awards, or if so should definitely exclude small awards and no cost extension (“Active”) awards [which no longer hit the NIH budget / cash flow]. Indeed, I’d argue it should be a cap on all NIH grants & contracts, not JUST RPGs.
But Drug Monkey is right that a huge problem with using a total $$$ cap (as opposed to one on the number of awards, ideally, capping number of all substantive grants & contracts) is that the NIH must maintain a broad and diversified portfolio, and different types of essential research have different costs. Unfortunately, a sub-text in all of this is that analysis of the changes in portfolio allocation supports the view that in the past decade, a substantial shift in portfolio allocation has preferentially steered resources into a set of “resource-building” exercises at much higher unit cost. If limited, or if taking place when the budget is growing, big bets on “bigger science” [GWAS ! Encyclopedias of (Cancer, Brain, etc) Genomes / Exomes ! ] can be great, but the resource shift is now destroying the diversity of research base. The same problem now challenges the long-time assumption that decisions should be made independent from selecting awards on the basis of lower F&A costs. Really, things need to move toward selections biased against higher costs, i.e., within a reasonable tier of indistinguishable percentiles, with a careful eye on program fit and NIH-wide ‘uniqueness’ (not just within one IC making that one fund /not fund decision).
Some of the early comments note use of indirect costs for building maintenance, etc, as if this were typical. If that is true, my institution is quite atypical, as I get billed for routine facilities repairs (leaky faucets, etc.) I think this kind of thing SHOULD be covered by indirect costs. On the other hand, a significant portion of indirect costs gets distributed to vice chancellors, deans, and departments (but none to individual faculty members). Some of the IDC funding goes to the library; a limited amount provides seed funding directed specifically to developing P01s (program project grants). At the departmental level, funding from other sources is so tight that the indirect costs that make it that far down the totem pole are simply absorbed into the general budget. There seems to be no general across-universities use of these funds.
I wonder if perhaps the issue of widely varied IDC rates, even among institutions in the same city, could be addressed in the same way that the federal government manages travel per diems: Simply specify a rate for each city/region, bearing in mind operating costs for that locale, and allow no variation from that rate. This will provide some degree of control over the variability in IDC rates among institutions.
RockTalkBlogStaff-
There is a good deal of talk about a “lost generation” of scientists. How about an update on the % of RPG/ % of $$ by age of PI graph that appeared in one of the NAS/NRC reports from ~8 yrs ago?
Ah yes, this is the graph I was remembering. The next page has the infamous age-of-first-R01 graph. A lot of things have changed since 2002…can we get an update on these numbers? It would go a long way to identifying the onset of any “lost generation” effects.
The key question is how will any of the changes impact the quality of science and productivity / innovation of the nation – this should be the first priority. I don’t believe changes should be made for other reasons.
In high school and college sports, there’s a vaguely similar situation in which small town schools can’t possibly compete with the big city schools (except in Hoosiers). So they created “divisions” based on school size.
You can imagine a similar method for this situation. One could create the divisions based on:
* school size or
* endowment size or
* total school research funding $$ or
* total investigator $$$ or
*** indirects (!).
No doubt, a combination of strategies would be better than just this one alone. But “divisions” could be a way of reducing some of the “rich get richer” problem.
In extracurricular sports, “competition” and broad participation is the goal and desired outcome. In NIH funded science, the goal is scientific production and advance.
Not simply to provide every possible University or college professor with an “opportunity” to do funded science.
I fear this flawed analogy is a common perspective and it echoes the Noonan position expressed previously on this blog that once a scientist is in the game, he more or less deserves his “slice of the pie” forever.
My reading of Rick’s analogy is different. It is, most of all, to provide all Universities and their scientists with the opportunity to “compete” for public funds in the “scientific production and advance” goals. True competition should eliminate Noonan’s position that “once a scientist is in the game”, he deserves “his slice of the pie” forever.
How about combining direct costs with indirect costs and making the entire award to PI. The PI should in turn negotiate the indirect costs with her institution and, if he feels the indirect costs are too high, she can relocate to another institution with a more favorable rate. This strategy would have institutions compete with each other for PIs and those that have lower indirect costs should merge as the winners. One would hope.
It would seem very logical to me as a study section reviewer and scientist receiving funding for 15 years that the size of awards could be trimmed some for modular RPGs but mainly for those exceeding $500,000 in directs per year. Moreover, the total number of grants per PI, per se, should be reduced WITH universities “picking up the slack” wherever possible in all of the above cases. Finally, exorbitant indirect costs to some of the more ‘entitled’ institutions should be renegotiated to a considerably lower rate. There is no reason, for instance, that a firmly endowed institution like Harvard should receive much higher indirects than the less well positioned. These funds could be used to fund many new investigator and other grants at varied institutions with equally-talented individuals.
I am in favor of the following steps:
o Reduce number of large payout grants, i.e., those in excess of $1 million. The scientific community needs to learn to do more with less. We are better off distributing funds to
a larger pool of qualified investigators rather than piling up money on a lucky few.
o Reduce PI salaries. Universities/Institutions should be paying the majority of salaries. NIH will get more bang for its buck by paying grad student and postdoc salaries. Paying large PI salaries is not an effective use of limited government funds.
o Re-negotiate indirect cost rates. The government is currently getting ripped off here. The claim was made above that the majority of F&A costs are used for space, lighting, heating, etc. Baloney. If you calculate the space and resource usage needed to conduct research for a $1M grant it comes no where close to $500k, which is at the lower end of what most universities charge. Good office space in most cities can easily be acquired for $25/sqft/year. Now be generous and allow for a 2000 sq ft. lab. That comes to $50k/year, or $250k for a five year grant period. And this all assumes that this grant is the sole source of income for the lab, which it probably isn’t. So universities have alot to answer for in charging these outrageous amounts.
Mostly they are wasting precious government resources on bloated administrative costs.
Where did you come up with $25/sq foot? You said this is for office space. Well, lab space is certainly more expensive, and it must be furnished/outfitted for the project. There is also the fact that numerous personnel are involved in administering the grant – ordering supplies, keeping records, overseeing compliance, etc. These costs add up quickly. That said, I would like to see administrations justify their indirect cost rates….
Reduce PI salaries. Universities/Institutions should be paying the majority of salaries.
I agree that at leat 50% of PI’s salary should be covered by universities/Institutions no matter public or private!!!!!!!!!!!!!!!!!!!!!!!!!!
I agree that a re-negotiation with Universities on the indirect costs is definitely needed. But on the NIH’s part the following things can also be done for better distribution of limited funds in the present times:
1. Introduce a referral system for PIs submitting RO1 grants whereby students and postdocs can grade their PIs on their mentoring abilities. This will not only ensure that the funding goes ONLY to the most deserving and qualified mentors but will also encourage better mentoring of a future generation of scientists.
2. Have a stricter policy on international travel for PIs and curb the amount of money spent behind it. Many PIs with large grants are constantly traveling to exotic locations to meet up with ‘collaborators’. In today’s age of communication, this can definitely be minimized.
Indirect costs should be a constant for all institutions. This would reward efficiency and avoid abuses of the system.
The NIGMS system of having a soft cap on overall grant support from all sources is a good model. It is supported by their survey of research productivity for labs as a function of support. Yet it allows exceptions for particularly compelling new grant applications from well funded investigators, but holds these new applications to a higher standard.
I think we need to focus on challenges at a higher level than cutting indirect costs. As a post-doc, you see every day that there are too many people in the system-unless the NIH budget grows by 2-3x. We need to start pushing the knowledge out into the world.
How can we drive more biological knowledge and technology into inventions/services which do good or create value and take pressure away from the academic system?
What about making it easier for academic scientists to pursue their ideas in the private sector? What about giving a series of “soft-landing grants”, grants that would fund a return to research life after founding a for-profit or non-profit venture? After so many years of training, it is very difficult for professors of even post-docs to leave academic jobs and try to push a research project out into the world. If it fails, it is hard to come back. By the time, we have the training to know biology deeply, people often have young families.
The NIH should award 20-100 soft-landing grants that would fund a return to academic life for a researcher who wants to push a technology out into the world. These could be prestigious-like pioneer awards-or work at the RO1 level-just like HHMI, funds a return to normalcy for un-renewed investigators.
In the short-term, this would require funds. I believe-in the medium term-it would create opportunities for more people with advanced training in the life-sciences to follow and work in for profit or non-profit ventures that are pushing our work out into the broader world.
I was shocked that PIs have to spend all of their money before the deadline, because they are afraid of the negative implication of not having spent it all. The result is to buy a bunch of things that they don’t need at the last minute in order to appear to have needed it all. This is wasteful. What if saving money became a positive incentive instead of a negative one? How about offer a guaranteed year extension with bonus, added bonus points on the next grant application, a larger percentage cut for the school, a boost to a higher tier for grant consideration for those who cut their cost by 40% and requested an extension? I see a lot of waste that could be utilized better by multiple scientists. I think that there are ethical implications that also need an overhaul. Because grants are so competitive, this may just be the time to tackle the beast and reform the system.
Here are a couple of more ideas…
1) How about offering incentives to decrease time of training graduate students. I believe that it can be done in 3.5-4.5 years, but many PIs are stretching it out because they don’t have the time and schools will cover it.
2) Increase incentives (points on grants) to train graduate students over holding onto higher paid postdocs for too many years.
3) Tighten the rules on expenditures and offer reward for whistleblowing!
4) Don’t fund scientist, institutions or organizations that with big ethical violations and problems. Make ethics a category for grant competitions.
What concerns me most viewing the data and the comments posted thus far is that any plan to deal with diminishing NIH funding will ultimately be pitting established, well-funded investigators against new investigators, using the well worn arguments of quality versus quantity and diversity.
Fact:
20% of the PIs get 50% of the funds.
Fact:
We can decide to keep funding the scientists that receive most of the funding under the assumption this maximizes “quality” or we can decide to redistribute funding to optimize quantity and diversity and find out how it affects quality.
Given the evidence that startup companies have provided most of the job growth in the US over the last decade, not to mention most of the creativity and innovation as well, I for one feel it is time to embrace quantity and diversity in NIH funding as I am confident it will only enhance quality and innovation in the end.
The American Society for Biochemistry and Molecular Biology thanks the National Institutes of Health for engaging the extramural research community in the discussion of NIH resources in financially constrained times. ASBMB is committed to preserving funding for basic research even in fiscally austerity. While we have submitted our response to the NIH email address, the public can view our response at the ASBMB Advocacy website.
Thank you again for your consideration.
Many senior investigators are getting salary or other support from more than 4 federal or non federal grants. Any Basic science investigator drawing support from more than 4 grants with <25% effort in not doing justice to any of these projects (sighltly differet for busy clinical investigators those who have only restricated time for reserach after clinical duties) . Most of these senior investigator are just enjoying "honarary postion" and junior investigators are forced to include them to use the fame of those seniors to increase success rate of their grant proposal in NIH review review/scoring process. NIH must restrict Principal investigator or co-investigatorship per investigator. A cap of 4 active grants in a fical year with average 25% effort or 3 months FTE can be implemented to prevent such "honarary position scam".
A cap of 3 active R01 sounds reasonable for basic scientists and a cap of 1 active R01 for clinical investigators. Clinicians ususally do not have enough effort/time to do the basic research and to mentor students/postdocs with limited productivity. It does not make any sense to me.
Indirect cost should be reduced at least 20% overall to save the money to support young investigators.
the solution is simple…. but not mentioned in any of the comments above:
mandatory retirement or a max age limit for PIs.
brilliant older scientists would then be expected to collaborate with younger PI as co-investigators
Over the past 20 years we have built an unsustainable system built on the need for more and more dollars to fund our factories that “grow” more and more researchers who cannot find jobs so they languish for years as post-doctoral fellows.
The system is seriously flawed as the best scientists are pulled from the laboratory to spend more and more time being “politicians” instead of scientists and training more and more students/fellows who cannot get jobs. Further, any system that encourages post-doctoral fellowships of more than 2 or 3 years with any single specific primary mentor is fatally flawed. If a mentor cannot complete the training of a post-doc in a few years, do we really think that we will impart knowledge more efficiently in the next 6 years?
With dollars growing ever more tight, I suggest that we need to create strictly enforced rules that limit each PI of federal grants to no more than a few (say 5) students and post-docs supported in whole or in part via federal funds at a time. Teeth should be added that will eliminate all federal research funding for departments/labs/PIs who try to circumvent this rule. This will break up the oligarchies that have developed across science and encourage PIs to stop being managers and to return to what we generally do best – science. It will also allow scientists to focus more on mentoring the students/fellows working with them so that they can impart the lessons to be learned in a period of no more than 2 or 3 years.
This will free up enormous sums of money, much more than the salary limits now being enforced. These funds can then be provided to younger independent scientists with fresh ideas and the energy to engage in science and not become managers.
The biggest drain on NIH resources stems mainly from issues related to salary support. Institutions no longer want to pay for their faculty (even though their faculty have responsibilities beyond their research) and push them to obtain, often times, 100% of their salaries from NIH research grants. I understand that for many extramural NIH researchers their main functions on campus are to do research, but placing large chunks, or all, of those salaries including fringe, on 250K annual budgets is a significant drain on monies that actually get directed to do the research. Funding is extremely competitive, clocks are ticking extraordinarily fast, and larger personnel costs to get the job done in an efficient manner are impeding the amount of actual dollars that go towards the research. Unfortunately, I think the only way to fix this will be to require institutes to make greater salary commitments to their faculty. (heck they own a portion of whatever IP we develop don’t they?)
Before thinking about how to mange NIH funds, we should review what the aims of these funds are. It seems to me the primary aim is to produce knowledge that is useful for the health of Americans, build research capacity, and support people and institutions that produce that produce high quality information. From that perspective, capping the number of grants or amount of funding to an individual investigator so that other investigators can be funded is very short sighted. It is essentially saying we should limit the research capacity in the countries most productive labs. Productive labs would have to let go their staff and research faculty, which also hurts many people, and could cause long-term damage to the strongest research groups in the country.
If we are concerned about ensuring building capacity strategies should focus on that. Safeguard the K-award programs. Strengthen first-award programs. Put your policies where your aims are.
The one area where the NIH may want to create some fairness is in indirects, capping the highest rates. But there too, we should be cautious. Indirects typically do not fully cover the costs of research.
We are in a rough patch. We have been through rough patches before. Using the pressure of budgets to execute wise changes in policy and funding makes sense. But let’s not go overboard. Things will improve.
NIH needs major overhaul.
i) There should be NO federal salary support for PIs. In fact, it should be illegal for receiving institutions to reward PIs for receiving federal grants. Instead, advancement should be based solely on publications and other public information-generating activities. Rewarding PIs for obtaining grants is perverse and counterproductive. PIs feeding at the trough just to support excessive salaries. Yes, excessive.
2) In addition to limiting compensation to PIs, there should be limits on grant sizes and numbers. There are diminishing returns when giving one PI more and more money. You might not want to believe it, but the brightest waste money too especially when they have too much on their plate.
3) NIH has had it good… too good, and it’s time to reevaluate the relative value of medical research. Like our healthcare providers, people in medical-related research are excessively compensated (by taxpayer money, not private get-what-ever-you-can money). Frankly, even if not one more medical advance occurred, we are still going to exceed the carrying capacity of the planet in the near future. “Saving” mankind is over-rated.
The subject of this post is critically important and I thank Dr. Rockey for visiting the idea. It seems necessary to revisit the goals of NIH funding at some point. Absent this exercise, I have some practical suggestions.
My four suggestions below (a-d) is based on the following basic ideas:
1) Investigators in universities should serve a dual role: researcher and educator
2) NIH should not pay for project managers and too much money in one researcher’s hand simply turns that investigator into a project manager
3) Good biomedical research should get better with time, and in some cases become self-sustaining.
4) NIH should pay for the essentials of research operations, not (indirectly) for the several vice presidents and multiple deans salaries, and some of the extravagant and half empty buildings.
Based on these ideas, immediate action by the NIH could stem the departure of many young scientists and future faculty to the industry.
a) limit total salary support to less than 50% for all PIs
b) limit total grant support to less than 500K direct to all PIs
c) set up a graded scale for research support (1,2,3). Young investigators 50%, mid-career investigators 30%, senior investigators 20%.
d) limit indirect costs across the board to all institutions
I don’t suggest that industry is bad. However, a healthy and critical mass of educators in universities is needed in order to sustain our educational systems. So long as NIH does not focus on small research grants for early investigators, and the NIH allows universities to use the generic nature of indirect to shift funds internally and support numerous administrators, the problem is likely to get worse.
With regards to smaller support and targeted to early investigators, I have seen statistics showing a small bump in funding younger investigators. However, I venture to guess that this bump is an anomaly.
The main point is that the number of grants should not be the operative measure, rather it should be total dollars. Neither should the name of the PI under which the grants are funded. What should be the operative measure is amount of federal dollars that goes in to a PI’s lab. For example, here is the statistics I’d like to see before I am convinced that the bump for young investigators is real: Take all funds for early investigators (no ARRA). Add total dollars that is going directly to the PI’s lab or a commercial company (for research or supplies) – in other words, if the money is going to another lab with a senior investigator, discount that amount. Now, let’s look at the trend lines.
Spread the wealth!!!! Cut the size of awards, limit the number of awards (or rather the $$$ amount) per PI to a set number, cut the F&As.
1. So much better science will be done when lots of people (with a reasonable amount of money) are given the opportunity to push the limits of their field rather than a select few who have lots of money. I have a modest NIH grant and it teaches me fiscal responsibility….something many PIs need to learn.
2. Do PI’s really need multi-million dollar grants….unless it is really justifiable it hurts everyone else.
3. Greedy administration and F&As….need I say more…our institutions are become so top heavy it is absurd.
One final note….universities should not have faculty positions (such as my own) that are 100% research based and 100% of my salary needs to come from grants. In a time of fiscal responsibility faculty positions should be a mix of teaching and research. If you get the grants, you can buy yourself out of teaching….if not, then you fall back on teaching….and so the cycle ebbs and flows during your career (essentially a safety net when times get hard).
I propose a system where your potential funding is linked to your institution’s overhead. This means that applicants would be divided into “risk pools” each separated by the negotiated overhead of the institution. In this instance, I define “risk as the dollars per finding. Obviously, if the grant starts out with a higher dollar amount there should be an expectation that the research would yield more than a lower dollar project. An application from universities with the highest overhead would accordingly have the lowest success rate, because they come with the most risk. This would create an incentive to bend the cost curve downward, instead of always negotiating for higher overhead. In this manner, it would promote academic institutes to be better at sharing resources, cutting costs and increasing efficiencies (including energy usage). This proposal could realistically create 25% more grants, without reducing the size of the award to researchers and perhaps even accelerate the diffusion of technological training and advancement from the coasts to the center of the country where costs are lower.
One potential way to stretch funding might be to cut R01 grants to 4 years and use the savings to increase R21s by 3 years at 150k/year and perhaps more depending upon the ratio of #R01s/#R21s.