Improving American Science:
FUNDING BASED ON NEED:
Minimal Basic Research Funding (MBRF)
A SERIES
NOTE: This 3-part series is part of a larger work-in-progress, in which many proposals regarding improving American science are set forth
Part 1: The current situation
Part 2: The problem
Part 3: A solution
Proposal for "Minimal Basic Research Funding Level" By Marion Freistadt
Series abstract
In this series, I discuss some problems with the funding of biomedical research in America today and propose some solutions. After discussing the current situation in Part I of this series, I describe the problem in the second part of the series. The problem being addressed here may be summarized as low return on investment. Large amounts of money are spent on research, yet much money and human effort is wasted. Supplies and equipment are purchased on arbitrary budget cycles, rather than as needed. Moral is low among many qualified scientists and some leave research (49% in one survey [http://www.amazon.com/Leaving-Science-Occupational-Scientific-Careers/dp/0871546949]). After discussing the difficulties, I propose a novel solution, which I call Minimal Basic Research Funding. I propose that researchers establish a “self-help” fund from their own money. I present estimates suggesting that, with small annual donations, in 5 years, 100 awards may be distributed. The psychological benefit of this safety net may save the careers of many scientists. This proposal may be characterized as research funding based on research need, rather than the peer review.
Summary of Part 1
In Part 1 http://thinkoutsidetheicosahedron.blogspot.com/ (Dec. 27, 2009 post), I described the current situation with regard to career paths for American, academic biomedical basic researchers (AABBR). After about 10 years of graduate and post-graduate training, the researcher seeks to acquire grants, generally from NIH (http://www.nih.gov/ National Institutes of Health, the federal government) while holding an academic position. I described the grant application process. A primary difficulty is passing the “peer-review” process. Universities have a vested interested in researchers’ acquisition of grant funding because in addition to the face value of grants, high overheads are charged.
Summary of Part 2
In Part 2, I discuss 4 problems with the current situation: 1. Why should progress in science be linked to money grubbing? 2. Wasted time and money, 3. Wasted materials, 4. Discouragement of scientists. First, I argue that pursuit of science and pursue of funding are two distinct functions and should not be inextricably linked. Next, I delineate ways in which applying for grant funding waste time and money. For example, a new study estimates that the cost of preparing a rejected grant as $40,000 (http://www.informaworld.com/smpp/content~content=a909104450). Wasted material is a consequence of finite budget time periods. Discouragement is the human cost. Many researchers experience undue stress, professional dissatisfaction. This may lead to clinical depression and/or possibly suicide. I reference several blogs that document these difficulties.
Summary of Part 3
In Part 3, I propose the establishment of Minimal Basic Research Funding. Here I establish the concept, and propose several ways that this could come about. If a researcher meets an established standard of productivity, yet lacks success in fund raising, a minimal level of funding which I call “Minimal Basic Research Funding (MBRF)” should be available to that person to continue researching. My proposal differs from others in that I suggest that researchers themselves contribute small amounts toward this end, similar to an insurance plan. In this document, I address several details of the proposal such as the establishment of criteria for productivity. I next address quantitative details. I propose the establishment of a 501c3 organization for this purpose. Using just researcher-based donations, I propose that in 5 years, if 3000 people donate $300 per year, sufficient funds would be available for 100 awards. Other sources would enhance the fund.
Part 3: A solution: Minimal Basic Research Funding
The basic idea and how to achieve it.
I propose that if a researcher meets an established standard of productivity, yet lacks success in fund raising, a minimal level of funding which I call “Minimal Basic Research Funding (MBRF)” should be available to that person to continue researching. This funding would NOT be based on merit or peer-review.
My first purpose here is simply to state the principle: this funding should be available. Getting agreement about this principle is a first step. At the moment, since this idea is in its inception, there is no single, definitive way to envision that this may be achieved. Many new, good ideas start this way. However, the means should not be confused with the ends. One important goal of this document is to explicitly state the principle.
My second purpose here is to propose one possible mechanism for achieving this. I believe my proposal is unique. I propose that researchers themselves help pay for this, similar to a voluntary unemployment insurance policy. There certainly may be other ways to achieve MBRF.
(NOTE: I am not proposing to do away with peer-review. Peer-review is a best practice and much better than cronyism or favoritism. However, for the reasons outlined in Part 2 of this series, it needs improvement.)
In this document, then, I discuss the details of this proposal.
Issues discussed below are:
1. How to establish an acceptable standard of productivity?
2. What level of funding should be provided?
3. How many people would need or use MBRF?
4. Criteria other than productivity for use of the fund.
5. Where would this funding come from?
1. How to establish an acceptable standard of productivity?
As a starting point, I would suggest that submitted (or accepted) publications and submitted grant applications are evidence of productivity. Some type of “committee” (Board of Directors or reporting to them) would be established to set productivity criteria. This committee should be composed of BOTH funded and unfunded researchers. Committees may be local or national. These could be web-based committees. Clearly, criteria for productivity evolve.
2. What level of funding should be provided?
The level of funding would be set by Board of Directors and may change over time. Levels will also depend on the specific research area. MBRF should not be a substitute for long-term funding. Compared to regular funding, the awards should be very low. Here, as an example, I make a suggestion based on my field, at the time that I was in this situation. I believe that with $5000/year, productive research (carried out by the PI) could be continued until major awards are received. It is to be used only for supplies. No personnel costs would be provided.
The committee will also establish criteria to minimize abuse and maximize successful funding independence. Eligibility is obviated by any other funding. Also a "buddy" system to link veteran funded researchers with unfunded researchers may help.
3. How many people would need or use MBRF?
This is very difficult to answer, for many reasons. Just to get started, I make some wild estimates. I will give two polar extremes: an extremely high and a relatively low estimate.
A high estimate
In Part 1 of this series, I estimated that there are 25,000 unfunded US researchers seeking funding. However, this is probably not the number that will be eligible for MBRF. Several requirements (discussed below), such as that the person have no funding and that he or she is actively pursuing research as evidenced by submitted or accepted publications, may mitigate this number.
A low estimate
For another estimate of researchers eligible for MBRF, envision a single academic biomedical research department. Examining one department’s website (http://tulane.edu/sse/cmb/faculty.cfm#CP_JUMP_678514), out of 19 faculty, 3 have NIH funding (http://projectreporter.nih.gov/reporter.cfm) totaling roughly $2 million for FY 2009. Of those without NIH funding, 6 have published in the last 2 years (http://www.ncbi.nlm.nih.gov/sites/entrez). These are the “active” researchers. Lacking other data, I estimate that of these 6, half may have non-NIH funding (collaborations, other federal sources, such as NSF, other nonprofit funding agencies, or industrial sources). This leaves an estimated 3 researchers in this department requiring MBRF.
According the Department of Education (http://nces.ed.gov/programs/digest/d08/tables/dt08_230.asp), there are 270 public and not-for-profit institutions that are involved in research at either a “very high” or “high” or “Doctoral” level (including medical schools). (There are 131 AAMC accredited medical schools in the US [http://www.aamc.org/medicalschools.htm].) I estimate that each institution has an average of 4 biomedical research departments carrying out basic research (nonmedical schools probably have 1; medical schools may have 6). So, 4 times 270 equals 1080 relevant departments. 1080 departments times 3 eligible researchers per department equals 3,240 eligible researchers in the US.
So my estimates vary from 3,240 (times $5000 per year equals $16.2 million per year) to approximately 25,000 ($125 million/year) American researchers seeking MBRF.
4. Criteria other than productivity
In addition to meeting the established criterion of productivity, there are two other criteria to be eligible for an award. First, the person must have no current funding, not even as a collaborator. Second, the person must be actively seeking funding. The goal of this funding is to permit the person to continue researching until he or she establishes him- or herself as a self-supporting researcher.
5. Where would this funding come from?
I propose starting a 501c3 organization. The purpose of the organization would be to provide a mechanism to fund MBRF. Input funds would be placed in an endowment, which would generate interest. When the fund reaches a certain level, awards will be distributed from interest. Awards would be granted to institutions, to be used exclusively by the researcher for proposed research, in the same way that research funding is currently granted. The money in the fund should managed by professional, independent managers. Salaries for managers are not included in these estimates.
The numbers
Table I. Endowment amounts required to award specified number of annual awards. Calculated by the formula: A = Ae^.0725 - A , where A is Award amount. (http://mathforum.org/library/drmath/view/54632.html ).
| Number of annual awards | Total amount awarded | Endowment required |
| 1 | $5,000 | $66,489 |
| 10 | $50,000 | $664,894 |
| 50 | $250,000 | $3,324,468 |
| 100 | $500,000 | $6,648,936 |
| 500 | $2,500,000 | $33,244,681 |
| 1,000 | $5,000,000 | $66,489,362 |
| 3,000 | $15,000,000 | $199,468,085 |
Table I shows the endowment amounts required to distribute specified number of annual awards, assuming an interest rate of 7.25% (compounded continuously). Thus, in order to distribute one award, the fund would need an endowment of $66,489, while to award 1000 awards, $66,489,362 would be required.
I propose that researchers pay for this themselves, at least in part.
This would work like an insurance fund. Each researcher would contribute a small amount.
Table II shows two 5 year scenarios: 100 people donating $100/year and 1000 people donating $1000/year. In each case, cumulative funds with interest are shown.
Table II. Accumulated funds based on amounts and numbers of donations.
Interest compounded continuously, calculated by the formula : A= Pert, where P = Principal, r = interest rate, t = time in years and A amount after time t. (http://cs.selu.edu/~rbyrd/math/continuous/)
| | Added to Endowment that year | Cumulative donations | Cumulative with 7.25% interest |
| 100 donors, each donating $100/year |
| Year 1 | 10,000 | 10,000 | $10,752 |
| Year 2 | 10,000 | 20,000 | $23,121 |
| Year 3 | 10,000 | 30,000 | $37,289 |
| Year 4 | 10,000 | 40,000 | $53,457 |
| Year 5 | 10,000 | 50,000 | $71,846 |
| | | | |
| 1000 donors, each donating $1000/year |
| Year 1 | 1,000,000 | 1,000,000 | $1,075,193 |
| Year 2 | 1,000,000 | 2,000,000 | $2,312,079 |
| Year 3 | 1,000,000 | 3,000,000 | $3,728,896 |
| Year 4 | 1,000,000 | 4,000,000 | $5,345,710 |
| Year 5 | 1,000,000 | 5,000,000 | $7,184,586 |
Combining the information from the two tables, at the lowest level, 100 people contributing $100/year, in 5 years one award would be generated (based on Table I, endowment amount needed per award). This is a 1 award per 100 donors award rate. At this level, in 15 years the endowment would be $445,027, permitting close to 7 awards (data not shown). If 1000 people donated $1000/year, after 1 year, $1,075,193 would permit close to 15 awards. After 5 years at this level, $7,184,586 would permit 108 awards.
To fund 3000 researchers, $200 million would be needed. This could be reached in 10 years if 5000 people gave $2000/year.
This level of funding for a nonprofit, charitable organization is not extravagant. Endowments of the wealthiest 25 charitable foundations on March 4, 2008 (http://en.wikipedia.org/wiki/List_of_wealthiest_foundations ) range from $3.3 billion (Kresge) to $35.1 billion (Bill and Melinda Gates). In fact, compared to the overall annual NIH budget ($30 billion), $200 million is 0.7%. Considering that this could be saving or improving the careers of thousands of scientists, this is not expensive.
Obviously, getting 5000 people to donate $2000 per year for 10 years would be difficult. Other revenue sources would be required. However, I believe that this general approach, at a lower level, is feasible in the short term.
For the researcher-supported part of the fund, I propose an intermediate-sized, medium range target: 3000 people donate $300 per year ($25/month). This would achieve $6,466,127 in 5 years, permitting 100 awards.
Incentives for researchers to contribute:
1. It’s cheap. Researchers for whom this is relevant are presumably already employed, so they can usually afford a small amount such as $25-$100/month.
2. It would be 100% tax-deductible, so donations could automatically be deducted from payroll and included in the cafeteria plan.
3. A researcher must contribute to receive an award.
Several additional points must be made here. First, this is NOT a personal investment. Money donated is not returned and there is no “return on investment.” Like a church or synagogue, it is not a “fee for service.” You cannot only contribute when it’s time for your child’s bar mitzvah: the organization needs to run all the time. It is like insurance, in the sense that you hope you never have to make a claim. If the fund disbands completely, money should be returned to donors, with interest (less expenses).
Who will donate?
I believe that any researcher who has gone through the process of desperately needing grant awards and having all applications rejected, will be sympathetic to this cause. Given overall success rates of 25% and first time success rates of 10%, this is a significant proportion of the biomedical science population. Ultimately, most scientists may wish to donate, since they realize that they someday may face the same situation.
Other possible revenue sources
In the long term, other sources of revenue will be required. These may come from existing research funding mechanisms or from new fund-raising options.
NIH or other relevant federal agencies
The simplest solution would be if NIH or research institutions provided this funding. It has been shown (in Canada) that the cost of submitting rejected grant applications exceeds the cost of small grants themselves. (“Cost of the NSERC Science Grant Peer Review System Exceeds the Cost of Giving Every Qualified Researcher a Baseline Grant”
http://www.informaworld.com/smpp/content~content=a909104450
http://scienceblogs.com/clock/2009/04/why_eliminate_the_peer-review.php). Therefore, a theoretical source of funding would simply be abolition of submitting unfunded grant applications. However, if it was known which grants would be funded, the review process would be unnecessary.
I am not proposing that NIH fund MBRF, at least not now. If I were, this proposal would be no different from the multiple, existing cries for more money dedicated to research. While I support that, the essence of this proposal is different. Simply taking money from NIH would require that that money not be spent elsewhere in NIH.
Institutions
Some institutions provide “bridge” or interim funding http://www.bumc.bu.edu/medicine/research-in-the-department/announcement-of-bridge-funding-application-deadlines/ .However, such mechanisms are not universal. If such a policy became universal and easy to access, my goal would be accomplished.
Other fund-raising options—researcher initiated.
Here are a few other ideas: MBRF could be a required use of overhead funds. Royalties from research inventions that currently go to institutions could be used. Science-related art and music is another potential source of income.
HOW TO CHOOSE WHO GETS AWARDS?
Ideally, if a person contributes, that person should automatically get an award when you needed. However, initially, the fund will not be large enough do this.
Criteria for receiving awards will be set by the Board of Directors. A formula would be developed. It could reflect: order in which application was received, length of time without funding will be derived. Critical to the MBRF plan is that it would NOT be based on peer-review.
OBJECTIONS
This proposal will not solve all problems in American academic basic biomedical research. However, it is a modest proposal as a start to improving American science. Objections will certainly arise to this proposal, particularly about its practicality. This has been addressed throughout the proposal. (1) Another objection is the lack of peer-review. (2) Another objection is that this may be abused. (3) Finally, it may be objected that this money is TOO low: it is not enough to perform research.
(1) Undermines peer-review system
Although criticism has been raised against it (http://en.wikipedia.org/wiki/Peer_review), peer-review has proven to be a “best practice” for allocating funds. Its purpose is to identify work most deserving of funding. I do not object to peer-review; in fact, I support it. Ideally, applications are judged and funded purely on the basis of merit, in contrast to arbitrary or capricious ways. Traditional ways may be to fund one’s friends and deny funding to one’s enemies. Peer-review is certainly preferable to graft or kickback payments.
The current proposal is for a “tweaking,” a refinement of the process. It is intended for cases in which the peer-review process does not succeed in achieving the goal for which is it intended.
(2) Abuse: INCENTIVES TO AVOID ABUSE
First, this level of funding is very low, making the incentive to cheat very low. Moreover, many people who do not succeed at the NIH level, are able to procure funding from other sources. This is for people who do not succeed, even from other sources. However, I believe that the requirement to contribute may eliminate “free-loaders.”
(3) Award not high enough
The low level of funding may be considered a reason NOT to do this. It may be felt that the amount is too low to permit productivity. I have heard this raised, primarily by highly successful fund-raisers. Such individuals may be unfamiliar with the situation being discussed here.
Clearly the amount of funding must be evaluated regularly. Also, it will depend on the specifics of each researcher’s field.
Keeping the monetary level low is a strength of this proposal. First, this is meant for minimal survival. If this funding was at a higher level, it would not be the minimum required for sustainability. Moreover, higher level would not be feasible—it would take on the aspects of big money funding of science.
Conclusion
This document does not have all the answers. My purpose here is to publicize the basic concept of MBRF and that it could, at least in part, be initiated and funded by researchers. In addition, I propose a number of possible routes to achieving MBRF.
Advocacy groups for science funding:
http://www.aboutastra.org/about/index.asp
http://www.cnsfweb.org/
