Click for next page ( 81


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 80
4 Funding for Large-Scale Science Obtaining funding is an essential step in launching any scientific research project. For large-scale projects, the challenges encoun- tered in securing funding to pursue an idea are amplified and in many ways unique. Potential sources of funding include government agencies, philanthropies and other nonprofit organizations, and indus- try, each of which has its advantages and limitations. In the United States, the federal government has traditionally been the primary fonder of large-scale projects, as defined in this report, because of the high costs of such activities. Not surprisingly, however, the provision of federal funds for large- scale projects has frequently been controversial, both across and within scientific disciplines. The angst across disciplines stems from the sense that large-scale projects funnel an inequitable or unjustified portion of the funds available for science and technology in general to one particular field, thus shortchanging other fields and impeding progress toward use- ful advances. For example, this argument has been used in debates re- garding the proposal to build a superconducting super collider, which was eventually rejected, as well as the proposal for the international space station, which was narrowly passed. The tension within a field stems largely from disagreements over whether large projects or more tradi- tional small-scale projects are the most efficient, economical, and benefi- cial for moving a field forward in the long run. These questions were widely debated in regard to the Human Genome Project (HGP). Although the completion of the reference draft of the human genome sequence has been widely hailed as a major achievement that will greatly 80

OCR for page 80
FUNDING FOR LARGE-SCALE SCIENCE 81 advance the fields of biology and biomedical research, questions are still being raised as to what role, if any, large-scale projects should have in future biological research. Many believe that smaller conventional, hy- pothesis-driven projects initiated by individual investigators are the most effective way to advance the field. But given the success of the HOP, there is also great interest in launching similar projects aimed at producing databases and other research tools that could facilitate the progress and potential of smaller, independent projects. Indeed, as noted in Chapter 3, a number of such projects have already been initiated. Thus, perhaps the most relevant question now is not whether the federal government should fund large-scale biology projects, but what the appropriate balance is between funding for large- and small-scale science in biomedical research and how funding for large-scale projects should be allocated. Yet little effort has been made to reach a consensus on the latter question, either in the broad fields of biology and biomedical research or in the more fo- cused field of cancer research. Even if providing funds for large-scale science is now culturally ac- ceptable in biomedical research, questions remain as to whether NIH is structured to fund such research. There is no agreed-upon method for allocating funds to large-scale projects, and there are many obstacles to overcome in designating funding for such projects, in part because the procedures and mechanisms used to disburse funds are still based on the more traditional approach to science. For example, the current, conven- tional NIH peer review process for vetting most research proposals is not very favorable to large-scale projects, which may not be hypothesis driven and often have nontraditional goals. But such a vetting process is essen- tial for achieving credibility and buy-in by the scientific community. Knowledgeable members of the community must be able to evaluate ad- equately and fairly the importance of the large-scale research goals, the feasibility of the plan, the value of the end products, and the level of opportunity to move the field forward. Such evaluation is challenging within the confines of the current system in part because the nature, and thus the assessment, of the goals and deliverables of large-scale biomedi- cal projects are quite different than from those for the customary smaller projects. The organization and planning requirements for large-scale proj- ects are also more elaborate, and therefore likely to require additional oversight and interim endpoints to achieve long-term accountability. Meeting these requirements necessitates additional resources and efforts on the part of the fonder as well as the investigator. This chapter provides an overview of the funding sources and mecha- nisms available for scientific research, both in general and specifically for biomedical research, with special emphasis on issues that are most rel- evant to large-scale projects in biomedical research. The discussion begins

OCR for page 80
82 LARGE-SCALE BIOMEDICAL SCIENCE with a brief review of the history of and process for allocation of federal funds for scientific research. A detailed description of funding for NIH is then presented, followed by a discussion of nonfederal funding of large- scale biomedical research projects. Issues associated with international collaborations are also examined. HISTORY OF FEDERAL SUPPORT FOR SCIENTIFIC RESEARCH The U.S government has often used its monetary resources to pursue matters of national interest. As the country's foundations were being laid, scientific research was not a national priority because the nation relied less on matters of science than it does today. But although federal scien- tific pursuits had a slow start, strong foundations were formed in the early nineteenth century that made possible the significant momentum in government sponsorship of public-based scientific endeavors experienced in the early part of the twentieth century (see Appendix). While early government investment in scientific research programs focused on agri- culture, national security, exploration, and commerce, many private foun- dations, such as Carnegie, Rockefeller, and Smithsonian, were supporting a variety of university-based basic research projects. That dichotomy is no longer true, as the U.S. federal government now supports the majority of basic scientific research undertaken at the nation's universities. The earliest federal support for civilian research was authorized in the 1800s, and included large-scale projects such as the U.S. Coast Survey and the U.S. Geological Survey. However, these initial efforts did not support the scientific education, training, and basic research that is now the hall- mark of universities. The first federal support for basic research within universities was initiated by the creation of the Department of Agriculture and the Land Grant Colleges. A series of congressional acts, starting with the Morrill Act of 1862, provided the mechanism by which scientists at universities could propose research projects and obtain federal funding to carry them out. These developments played a substantial role in the forma- tion of a number of biological sciences in the United States, including bac- teriology, biochemistry, and genetics (Goldberg, 1995~. The creation of NIH eventually led to an analogous impact on biomedical research when it began providing federal funds for extramural projects at universities. Simi- larly, the creation of NCI in 1937 was instrumental in launching a federally sponsored campaign to understand and eliminate human cancer. The period of time surrounding World War II had a particularly sig- nificant impact on the government's investment in university-based sci- entific research and its willingness to underwrite big-science projects. During the decade from 1940 to 1950, several key events facilitated the creation or expansion of science-oriented agencies, such as the Office of

OCR for page 80
FUNDING FOR LARGE-SCALE SCIENCE 83 Naval Research, the National Science Foundation (NSF), and NIH, whose main objectives became the sponsorship of public research. A key initial impetus for the expansion of federally sponsored scientific research was Vannevar Bush's 1945 report to the President Science: The Endless Fron- tier but other leaders also played important roles in developing the cur- rent mechanisms for federal support of science, particularly with respect to the more applied fields of research. (For a detailed review, see Appen- dix.) The resultant changes ensured that federal funding for university- based scientific research would become the accepted and expected norm that it is today. These changes also paved the way for federal support of future big-science projects in such fields as high-energy physics, space science, and biology. ALLOCATION OF FEDERAL FUNDS FOR SCIENTIFIC RESEARCH The process for appropriating federal funds is both complex and treacherous. The separation of powers between the executive and legisla- tive branches of the U.S. government makes it difficult to ascribe respon- sibility for any particular government action. Decisions regarding bud- gets and funding priorities are made through complex procedures that are influenced by many factors and federal entities. Determining funding priorities in a fluctuating social and economic environment is difficult, and by its very nature controversial. The U.S. government must deter- mine how much money should be allocated for scientific research as a whole, and how to divide that money among the various claimants in the science and technology community (Green, 1995~. Yet broad priority set- ting is generally resisted by the recipients of federal funding because it orders the importance of research investments in ways that groups within the scientific community often do not support (Office of Technology As- sessment, 1991; McGeary and Merrill, 1999~. The process is inherently contentious because priority setting creates winners and losers. Although American science is unparalleled in its scale and scope compared with that of other nations, the publicly financed sector exists in an economy of scarcity because scientists and institutions will always have more ideas for research projects than can be funded (Greenberg, 2001~. In resisting priority setting, the scientific community aims to maintain high levels of funding for all fields, instead of risking cuts in any particular one. There are few established methods for comparing, evaluating, and ranking research programs regardless of their size, although criteria have been proposed (Office of Technology Assessment, 1991; see Box 4-1~. Even ~ Vannevar Bush made a strong distinction between basic and applied research, and gen- erally did not advocate government support of applied research.

OCR for page 80
84 LARGE-SCALE BIOMEDICAL SCIENCE

OCR for page 80
FUNDING FOR LARGE-SCALE SCIENCE 85 within a discipline, distribution of funds can be contentious, as demon- strated by the 1995 National Research Council (NRC) study that pro- duced the report Setting Priorities in Space Research: An Experiment in Meth- odology, in which no consensus was reached on how to make allocations. The challenges associated with allocating funds across scientific fields are even greater. No single organization looks across the federal research system to determine priorities, and there is currently no formal or explicit mechanism for evaluating the total research portfolio of the federal gov- ernment in terms of progress toward national objectives. Mechanisms that may help determine priorities include the individual agency advi- sory committees (see Box 4-2) and peer review procedures, the Office of Science and Technology Policy and other White House advisory com- mittees, and the NRC system. Even with these mechanisms in place, however, there is no way to avoid competition among the various claims on federal science funds or to balance the federal research portfolio systematically. As described in more detail below, a variety of unrelated agency budgets could be in competition for the funds available under the juris- diction of an individual appropriations committee, and no single sub- committee is responsible for all science funding agencies, making it very

OCR for page 80
86 LARGE-SCALE BIOMEDICAL SCIENCE difficult to prioritize across disciplines. The NRC (1995) identified this predicament as a major obstacle in allocating federal funds for science and technology equitably and appropriately across the various fields and agencies. The report recommended changes to the process that would allow presentation and examination of the entire, comprehensive science and technology budget before it is disaggregated among the various com- mittees and subcommittees. Only recently have Congress and the Admin- istration begun to discuss the balance of funding among fields. For the fiscal year (FY) 2001 budget cycle, the Bush Administration stated for the first time that balance would be an explicit criterion in developing its budget request. The budget contained a component called "Federal Sci- ence and Technology," which was meant to represent investment in new knowledge and know-how. This was a break from tradition, but still does not enable priority setting among fields (National Research Council, 2001a). Thus, the NRC report recommended that the executive branch and Congress institutionalize processes for conducting and acting on an

OCR for page 80
FUNDING FOR LARGE-SCALE SCIENCE 87 integrated analysis of the federal budget for research, by field as well as by agency, national purpose, and other perspectives. One ongoing change in budget allocations is the effort by the Office of Management and Budget (OMB) to apply stricter performance measures in funding federal research agencies based on the Government Perfor- mance and Results Act (GPRA) of 1993 (Hafner, 2002~. GPRA requires agencies to manage and budget according to performance standards as a way of promoting efficiency, accountability, and effectiveness in govern- ment spending. However, it is still unclear to what extent Congress will adopt more definitive guidelines, with an emphasis on output, for scien- tific research. In the past, Congress has been amenable to investing in undifferentiated science, with knowledge as the outcome. Indeed, GRPA has caused consternation among the research agencies because few have had any experience in actually measuring the results of their programs, and they are unaccustomed to the increased scrutiny. Many researchers have argued that the results of ongoing basic research cannot be bench- marked or measured (Lekowski, 1999~. A 1999 report addressing the issue of assessing research in compli- ance with GPRA agreed that basic research cannot be measured directly on an annual basis because its outcomes are unpredictable, and there is generally a significant time delay between the generation of new knowl- edge and its practical application (National Research Council, l999~. How- ever, the report did suggest that measures of quality, relevance, and lead- ership are sound indicators of eventual usefulness and can be reported regularly while research is in progress. The report also encouraged bench- marking of programs in one agency against other federal programs, as well as international benchmarking, as a measure for fostering quality and leadership in a given field of research. The report made two addi- tional major recommendations: that research programs also be graded on whether they perform an effective education and training function, and that interagency programs be graded according to how well they are coordinated. The FY 2003 federal budget marks the first year that OMB has actually linked management performance with research budget priorities (Softcheck, 2002~. The process for using the new performance criteria and standards for applied research and development (R&D) was piloted with the Department of Energy (DOE) (Hafner, 2002~. Standards for evaluating basic R&D are still in development, with plans to implement them in FY 2004 across all federal research agencies. Assessment parameters will be refined in consultation with a variety of scientific bodies, such as the National Academies' Commit- tee on Science, Engineering, and Public Policy (COSEPUP). As part of the new focus on performance, OMB recently released a red/yellow/green scorecard for each federal agency (with red being the

OCR for page 80
88 LARGE-SCALE BIOMEDICAL SCIENCE lowest score). Almost 80 percent of those reviewed received red scores in the five rating categories. Only one agency, NSF, received a green score in one of the five categories for financial management (Softcheck, 2002~. However, a recent follow-up study by COSEPUP also examined the ways in which federal agencies that support science and engineering research are responding to GPRA (National Academies, 2001~. The committee found that although there is significant variation in responses, NIH, NSF, the Department of Defense (DOD), DOE, and the National Aeronautics and Space Administration (NASA) have all taken steps to develop report- ing procedures to comply with GPRA requirements. The committee also concluded that some agencies were using GPRA to improve their opera- tions, but that oversight bodies needed clearer procedures to validate and verify the agency evaluations, and that communication between over- sight bodies and the agencies was not adequate. An overview of the process for appropriating and allocating federal funds in the United States is shown in Figure 4-1. Briefly, the President, in conjunction with OMB, submits a detailed budget that includes many line-item requests about 15 months prior to the start of the budget's fiscal year. OMB crafts the budgets of research programs to reflect the priorities of the President, and attempts to compare the projected costs, benefits, and risks of certain programs to set realistic targets for the budget. The President's budget is submitted to both the House and Senate budget committees. These two committees review the budget and make changes to broad funding areas, called functions, in the areas of health, defense, civilian R&D, and so on. Congressional authorizing committees2 then can either authorize or not authorize (as nearly occurred with the space sta- tion) the use of the funds by specific government agencies and programs. The revised budget is next given to the House and Senate full appropria- tions committees and is divided among the 13 corresponding appropria- tion subcommittees,3 which are mirrored on the House and Senate sides (see Table 4-1~. Although specific budget items may have been outlined by the President, the budget committees, the authorizing committees, and the appropriations committees have the decisive influence over the funds distributed to R&D agencies. Each of the 13 appropriations subcommittees from the House and Senate writes a bill that is submitted back to the respective full committee, and the bills are taken to the House or Senate floor. Once the bills have 2 Authorizing committees supervise the activities of agencies under their jurisdiction and pass laws (authorization bills) directing those activities and setting nonbinding ceilings for their budgets. 3 The appropriations committees set the actual budgets of all agencies in the government.

OCR for page 80
FUNDING FOR LARGE-SCALE SCIENCE House Budget Committee President's Budget Request - 1 ~ Senate Budget Committee Input from House authc committees 1 *13 House ~ | House Full | Appropriations ~ Appropriations Subcommittees ~ U~~ | House | | Senate | | Floor | | Floor | 1 1 *Subcommittees: Agriculture Commerce Defense District of Columbia Energy and Water Foreign Operations Homeland Security Interior Labor, HHS, and Education Legislative Military Construction Transportation, Treasury, and Independent Agencies VA, HUD, and Independent House Agencies Floor 1 89 1 1 1 3 Congressional Conference Committees Input from Senate authorizing ~ committees Senat'3 Full ~ *13 Senate Appropriations _ Appropriations Committee _ Subcommittees | Senate | L: 1 ~ Presidential Approval FIGURE 4-1 Federal budget approval process. been approved, they go to a congressional conference committee made up of House and Senate members from the corresponding appropriations subcommittees. The further revised individual bills, often a compromise between House and Senate versions, are taken back to the floor and sub- mitted for a vote. If approved, each bill goes back to the President for signing. As the President signs the final bills, they become laws. The "budget" for R&D is contained in the aggregate of appropriations bills passed for the year. One limitation of this system that may be especially relevant to the funding of large-scale research projects is that federal appropriations are

OCR for page 80
So LARGE-SCALE BIOMEDICAL SCIENCE TABLE 4-1 Selected Congressional Appropriations Committee Jurisdictions Committee Jurisdiction Appropriations Committee or Subcommittee Name Senate House Agriculture 1. Department of Agriculture (except Forest Service) 2. Farm Credit Administration 3. Commodity Futures Trading , - . . Commission 4. Food and Drug Administration (DHHS) 4. 7. 1. Adulteration of seeds, insect pests, and protection of birds and animals in forest reserves 2. Agriculture generally 3. Agricultural and industrial chemistry Agricultural colleges and experiment stations 5. Agricultural economics and research 6. Agricultural education extension services Agricultural production and marketing and stabilization of prices of agricultural products and commodities (not including distribution outside the United States) 8. Animal industry and diseases of animals 9. Crop insurance and soil conservation 10. Dairy industry 11. Entomology and plant quarantine 12. Extension of farm credit and farm security 13. Forestry in general, and forest reserves other than those created from the public domain 14. Human nutrition and home economics 15. Inspection of livestock and meat products 16. Plant industry, soils, and agricultural engineering 17. Rural electrification 18. Commodities exchanges 19. Rural development

OCR for page 80
FUNDING FOR LARGE-SCALE SCIENCE 119 2,500 - 2,000 - cn 1,500- Cal o . _ 1,000- 500 - /,,f'/ A../ /,W,,f 600 - /- 500 - (A 400- '~ /.f) ... 300- 200- Fox 100 - o / / / j,,f'k' Genomics firms Pharmaceutical and Government and biotechnology firms non-profit organizations U.S. government Foreign governments non-profits Foreign non-profits FIGURE 4-3 Worldwide funding for genomics research, 2000 (millions of SU.S.~. SOURCE: World Survey of Funding for Genomics and Stanford in Washington Program, http: / /www.stanford.edu/class/siwl98q/websites/genomics.

OCR for page 80
120 LARGE-SCALE BIOMEDICAL SCIENCE 80 - 70 - 60 - 50 - 40 - A. Number of genomics firms with publicly traded stock o #firms ................ ................. .................. .................. ................... ................. .................. 1994 1995 1996 1997 1998 1999 2000 8 10 14 19 25 28 73 Year 100 90 80 oh ~ 70 a) 0 60 . _ 5] .- 50 g 40 30 20 10 o B. Growth in market value of nenomics firms 1994 1996 1998 2000 Year FIGURE 4-4 Growth of commercial genomics. A: Number of firms with pub- licly traded stock. B: Growth in market value of genomics firms. SOURCE: World Survey of Funding for Genomics, Stanford in Washington Pro- gram, http: / /www.stanford.edu/class/siwl98q/websites/genomics.

OCR for page 80
FUNDING FOR LARGE-SCALE SCIENCE 121 large-scale projects undertaken solely in the private sector, as is discussed further in Chapter 7. These very concerns recently led to a unique public-private collabo- ration to sequence the mouse genome. NHGRI began a mouse sequencing project in 1999 by providing funding to 10 laboratories using a combina- tion of sequencing strategies, such as sequencing randomly chosen DNA or particular DNA regions of biological interest. In the spring of 2000, the publicly funded group chose to a adopt a hybrid strategy combining data generated by the whole-genome shotgun approach for most of the genome with some sequences generated the more traditional way, using genomic maps. This decision was based on the success of the Drosophila sequencing project26 and on pilot projects conducted by the mouse se- quencers (Pennisi,2000b). Shortly thereafter, Celera began sequencing the genomes of three different strains of laboratory mice on its own. Within 6 months, Celera was offering access to a database of these sequences to anyone willing and able to pay a user fee. Because of a strong desire at NIH and in the research community to have a sequence that was freely available to the public, a new public-private consortium was announced in the fall of 2000, with the goal of sequencing the genome of a fourth mouse strain (Marshall, 2000~. Six Institutes at NIH, including NCI, two companies, and two nonprofit organizations provided $58 million to se- quence the genome in 6 months using the whole-genome shotgun ap- proach employed by Celera. The new money was divided among only three sequencing centers two in the United States and one in the United Kingdom to complete the work. On May 6, 2002, the Mouse Genome Sequencing Consortium announced the completion of a draft sequence for one common laboratory strain of mouse, which is available free of charge through the Internet (Marshall, 2002b). In fact, the consortium released data in real time to a public database throughout the project, with no restrictions. However, the public project was criticized initially for not making a greater effort to assemble the mouse genome sequences into a form that would enable the study of gene structure and function (Marshall, 2001~. A less competitive approach was subsequently taken in sequencing the rat genome through a public-private consortium. That project, which is also using a strategy that combines a map-based sequencing approach and Celera's whole-genome shotgun approach, is funded jointly by NHGRI and the National Heart, Lung, and Blood Institute (NHLBI) (Marshall, 2001; Hafner, 2001~. In this case, however, a substantial fraction ($21 million out 26 In the case of the Drosophila genome, a group of NHGRI-funded researchers supplied Celera with more than 10,000 cloned fragments of DNA to which the company applied the shotgun sequencing method. The data were released to the public (Pennisi, 1999~.

OCR for page 80
22 LARGE-SCALE BIOMEDICAL SCIENCE of a $58 million total) of the most recent batch of NIH funding will go to Celera to perform the sequencing. Much of the remaining funding will go to a second sequencing company, Genome Therapeutics Corporation. Be- cause the funding is derived from federal sources, the participants agreed to abide by a set of mandatory data-release rules that require grantees to publicly release raw sequence data on a weekly basis. This approach may be a model for future endeavors. While avoiding duplication of public and private efforts, it provides a cost-effective mechanism for producing a pub- lic good (a freely available sequence database) using industry standards for staffing, management, and quality control. Another approach to establishing public-private collaborations is a cooperative research and development agreement (CRADA). Under the Federal Technology Transfer Act (FTTA) of 1986, federal agencies have been mandated to encourage and facilitate collaboration among federal laboratories, state and local governments, universities, and the private sector in order to assist in the transfer of federal technology to the market place. One vehicle for this collaboration is through a CRADA. Examples of products that have resulted in part through a CRADA include Havrix~ and Taxol~. A CRADA is a contractual agreements between one or more federal laboratories and one or more industrial or university partners, under which the federal laboratories provide personnel, services, facilities, equipment, or other resources with or without reimbursement and the nonfederal par- ties provide funds, personnel, services, facilities, equipment, or other re- sources toward the conduct of a particular R&D program. The purpose of a CRADA is to make available government facilities, intellectual property, and expertise for collaborative interactions aimed at developing useful, marketable products that would benefit the public. The terms of a CRADA are usually brief and flexible so that each agreement can be negotiated and tailored to the needs and resources of the participating parties. There must be an intellectual contribution, which may take the form of materials, in- strumentation, or expertise, from all parties to the agreement, but the fed- eral government does not provide funding to nonfederal parties. However, a major benefit to an industrial collaborator is that it may obtain a first option for licensing of patents that result from the CRADA. This type of agreement was recently used to establish a joint project between DOE and two companiesCelera and Compaq to develop the next generation of software and computer hardware tools for computa- tional biology (Washington Fax, lanuary 29, 2001~. Such bioinformatics tools 27 See , and NIH Office of Technology Trans- fer, .

OCR for page 80
FUNDING FOR LARGE-SCALE SCIENCE 123 are necessary to process data from large-scale projects such as the HOP, structural genomics, and proteomics. DOE will provide $10 million for work at Sandia National Laboratories. The exact financial contributions from the two firms have not been disclosed, but are also probably in the multimillion dollar range. Compaq and Sandia will work together on de- veloping system hardware and software, while Celera and Sandia will col- laborate on new visualization technologies for analyzing the massive quan- tities of experimental data generated by high-throughput instruments. Nonprofit Funding of Large-Scale Biomedical Research Nonprofit organizations, while making a small funding contribution in comparison with private industry and the government, have also played an important role in genomics research and could potentially con- tribute to other large-scale biology projects. Nonprofit28 organizations come in a variety of different forms, including volunteer organizations, such as the American Cancer Society, that continually raise money to support research; endowed philanthropies, such as HHMI29; and even organizations set up by for-profit companies, such as the SNP Consor- tium. Examples of science-funding philanthropies are listed in Table 4-4. Profits generated by the bull stock market of the 1990s fueled unprec- edented growth in philanthropic foundation assets and giving. In 1998, grant-making nonprofits spent more than $1 billion on science, but the recent downturn of the U.S. stock market has quelled that growth. As noted earlier, philanthropies such as the Carnegie and Rockefeller Foundations played a leading role in funding and shaping basic science in the United States before World War II and by doing so even gave rise to new fields, such as molecular biology. Many organizations try to continue that tradition today by focusing on filling perceived gaps in federal funding and by defining highly specific targets for research (Cohen, 1999~. In some ways, nonprofits have an advantage over government funding in their ability to change course quickly and to pursue nontraditional or high-risk projects. They often undertake peer review in a form much different from that of NIH, and some ignore the peer review process altogether. Many also have less-stringent reporting requirements with respect to progress and outcomes than does the federal government. While these characteristics may be considered risky at the very least, they certainly facilitate the fund- 28A nonprofit organization must spend 5 percent of its assets each year or face tax penal- ties. 29 Because HHMI hires researchers as employees instead of awarding grants, it is in a different category and has to spend only 3.5 percent of its assets annually.

OCR for page 80
124 LARGE-SCALE BIOMEDICAL SCIENCE TABLE 4-4 Selected Science-Funding Philanthropies 1999* 1999* Science Name Founded Assets Expenses Research Focus Wellcome Trust 1936 $19.2B $640M Biomedical, no cancer Bill and Melinda 1994 $17.1B $230M Vaccines, reproductive Gates Foundation medicine, public health David and Lucite 1964 $13.5B $84.7M Ocean sciences, computer Packard science, math, natural Foundation science, engineering, interdisciplinary Howard Hughes 1953 $12B $427.7M Biomedical Medical Research Institute Pew Charitable 1948-79 $4.7B $6.95M Biomedical, neuroscience Trusts Rockefeller 1913 $3.5B $20M Reproductive health, Foundation agriculture, vaccines, epidemiology, malaria Andrew W. Mellon 1940-69 $3.5B $3.1M Contraception, repro- Foundation ductive biology, ecology Kresge Foundation 1924 $2.1B $4.6M Scientific equipment Carnegie 1911 $1.7B $1M Russian science Corporation W. M. Keck 1954 $1.7B $38.M1 Science, engineering, Foundation medical, astronomy Donald Reynolds 1954 $1.4B $35.2M Cardiovascular clinical Foundation over 5 research, geriatrics years Doris Duke 1997 $1.4B $13.8M Physician-scientists, no Charitable Trust animal research Alfred P. Sloan 1934 $1.2B $5.6M Astronomy, molecular Foundation evolution, neurobiology, marine biology, compu- tational biology Burroughs 1955 $669M $35M Biomedical Wellcome Fund Edna McConnell 1969 $640M $898,000 Trachoma, onchocerciasis Clark Foundation vaccine Welch Foundation 1954 $362M $23M Chemistry, primarily in Texas Carnegie Institution 1902 $527.1M $31.4M Astronomy, geophysics, of Washington plant biology, embryology M. J. Murdock 1975 $525M $4M Natural sciences, primarily Charitable Trust in Pacific NW James S. McDonnell 1950 $480M $19M Neuroscience, genetics, Foundation astronomy, complex systems Arnold and Mabel 1977 $450M NA Chemistry, biochemistry, Beckman medicine Foundation

OCR for page 80
FUNDING FOR LARGE-SCALE SCIENCE TABLE 4-4 continued 125 Name Founded 1999* Assets 1999* Science Expenses Research Focus Whitaker Foundation Charles A. Dana Foundation Research Corporation Camille and Henry Dreyfus Foundation Ellison Medical Foundation 1975 1950 1912 1946 1998 $390M $311M $152.3M $125M N/A $65.7M Biomedical engineering $10M $6.4M $3.4M 100M over Aging 5 years Neurosciences Chemistry, physics, astronomy Chemistry *Many of these are estimates. SOURCE: Cohen (1999~. ing of unconventional or controversial projects. With the exception of the largest organizations, such as HHMI, the Wellcome Trust,30 and the Gates Foundation, however, single-handedly funding a large-scale initiative or providing long-term support beyond pilot projects may not be feasible. A joint venture is a possibility, but philanthropies often find it unpalatable to work together or with the federal government, fearing that they will dilute their own impact and identity (Cohen, 1999~. Such was not the case, how- ever, for the Wellcome Trust, which contributed heavily to several recent large-scale projects, including the internationally funded HGP. In most cases, investigators look to federal funding sources to continue a project that was launched successfully in a pilot or proof-of-principle stage using philanthropic sources. Such grant applications may then be viewed as less risky, but investigators may still encounter difficulties in obtaining NIH funds if the projects are very costly and the applications have not been solicited through a PA or RFA. ISSUES ASSOCIATED WITH INTERNATIONAL COLLABORATIONS The drive to achieve international standing and recognition in a par- ticular field can promote competition and impede scientific cooperation. Nonetheless, the international collaborative approach for scientific re- 30The Wellcome Trust outspends the combined budgets of the United Kingdom's main government funders of biological research. Wellcome targets specific diseases, but avoids those that are relatively well funded (including cancer).

OCR for page 80
26 LARGE-SCALE BIOMEDICAL SCIENCE search has become commonplace for large-scale projects in such fields as high-energy physics, which require very large and expensive facilities. These collaborations may still be contentious because of competition among research groups or nations, but the end products of the research generally do not have direct commercial value. In the case of molecular biology and biomedical research, however, international competition is exacerbated by the fact that patents on new discoveries can be extremely lucrative. The lure of potential profits and market shares adds an addi- tional level of complexity to negotiations for collaborative projects. These challenges are intensified by basic difficulties in organizing and manag- ing projects undertaken on a global scale. Establishing uniform priorities and goals for the overall project and for each participant is highly prob- lematic and is complicated by difficulties in communication across cul- tures, languages, and political environments. Nonetheless, the scientific and engineering communities in the United States benefit from ideas and technologies developed around the world, and participating in international scientific and technical collaborations and exchanges may provide unique opportunities for addressing major problems or questions. Indeed, a 1995 NRC report recommends that the United States should pursue international cooperation to share costs, to tap into the world's best science and technology, and to meet national goals (National Research Council, 1995~. The World Health Organization has led the way in creating structures to enable international cooperation for health R&D as a tool for economic and social development. According to the Global Forum for Health Research, the international activities bud- get for NIH increased steadily from 1991 to more than $200 million in 1998. There are international programs within the various NIH Institutes, but a breakdown of these activities was not available to the committee, and it is unclear how much of that funding went toward projects that would qualify as large-scale research as defined in this report. SUMMARY It is difficult, if not impossible, to quantify the total amount or pro- portion of biomedical research funding that is spent on large-scale re- search projects, primarily because of variation in definitions and report- ing practices. As examples described here and in Chapter 3 clearly indicate, however, large-scale science projects are certainly being under- taken with funding from federal as well as nonfederal sources (the latter including industry and philanthropies and other nonprofits). The objec- tives and cultures of these different sources may vary considerably, yet partnerships among diverse funding sources could offer unique opportu- nities for undertaking large-scale endeavors if the challenges entailed can

OCR for page 80
FUNDING FOR LARGE-SCALE SCIENCE 127 be overcome. In particular, public-private collaborations provide a way to share the costs and risks, as well as the benefits, of such efforts. Interna- tional collaborations may present the greatest challenge of all, but also offer potentially unique opportunities. Some of the challenges involved, such as organization and management of projects and concerns about intellectual property, are covered in more detail in Chapters 5 and 7. Federal funding for large-scale science projects continues to be con- troversial. Proposals for undertaking such projects often generate criti- cism and debate, both across and within fields. Although this debate on the relative value of such projects is crucial to their success, resolving these arguments is complicated by the fact that there is no consistent, established way to balance the allocation of funds across the various dis- ciplines, or across big versus small projects. Over the course of the last century, however, scientists have come to expect federal funding for re- search, and those pursuing large-scale projects are no exception. Further- more, former acting NIH director Ruth Kirschstein has noted that while the "bedrock" of the agency's research will continue to be individual investigator-initiated inquiry, the nature of scientific investigation is changing such that current research questions are more likely to require the efforts of multidisciplinary teams working with expensive instruments in specialized facilities (Haley, 2001~. Similarly, current NIH director Elias Zerhouni has remarked that the model of the traditional NIH grant "will evolve into different shapes because multidisciplinary science requires collaborations." But he has also noted that "at the end of the day you also need [principal investigators] who themselves have an inherent under- standing of [multiple] fields so they can ask the right questions" (Kaiser, 2002:1~. According to Lake and Hood (2001), one of the outstanding chal- lenges for contemporary biology is the integration of hypothesis-driven science with a new discovery approach to science that is, defining all the elements of a biological system as a key information resource, and study- ing the entire system rather than asking questions about highly specific components. The examples described in Chapter 3 indicate that there is flexibility within the NIH procedures that allows for some large-scale research en- deavors. Within NIH, however, recent funding patterns suggest that per- haps only the Institutes with the largest budgets (e.g., NCI, NIGMS, and NHLBI) can independently handle the launch and support of a large- scale research project. Others may not have enough funds or flexibility in their budgets. For the smaller Institutes, undertaking such projects may require action and support on the part of the NIH director, or at least collaborative efforts among smaller and larger Institutes. NHGRI may be an exception to this generalization, since it was created specifically to undertake the large-scale HOP.

OCR for page 80
28 LARGE-SCALE BIOMEDICAL SCIENCE Some currently available funding mechanisms at NIH are amenable to large-scale projects and have already been used for such projects. Most of these efforts depended upon the solicitation of applications through PAs or RFAs that were issued for a specific topic of research. Unsolicited proposals for large-scale projects face what may be insurmountable ob- stacles in the form of grant size restrictions, traditional peer review expec- tations, and yearly fluctuations in the congressional allocations to NIH Institutes and Centers. Furthermore, using the RO1 funding mechanism (the most common for unsolicited grants) for large-scale projects could lead to greater competition in the short term between scientists conduct- ing large-scale and small-scale biomedical research because, absent a net increase in funding, each multimillion dollar grant would proportionally reduce the number of traditionally sized ROls awarded. As NIH ap- proaches the completion of the budget doubling of recent years, there is already concern that the percentage of new applications funded will drop because of commitments made during the growth years (Korn et al., 2002; Jenkins, 2003b). At any given time, approximately 70 percent of the Insti- tutes' funds are allocated for noncompeting renewals of awards made in previous years. How are decisions to be made regarding the types of projects to be undertaken and the most pressing needs of the field? If NIH wishes to facilitate the process of funding large-scale projects that generate data- bases and other research tools, it may be helpful to change, or in some cases standardize, the decision-making procedures within the Institutes and Centers. For example, the traditional peer review process favor proj- ects that are hypothesis driven. To date, in fact, none of the large projects funded by NCI have been reviewed through the CSR.3~ According to Craig Venter, the traditional dogmatic approach to peer review denies that biology is descriptive and impedes the progress of discovery (Lewis, 2001~. While no one would deny the value of hypothesis-driven research, balancing the research portfolio with multiple approaches could enhance the progress of science overall. Changes in the peer review process could provide a first step in achieving that balance. A critical assessment and standardization of the procedures for issuing PAs and RFAs would also be useful for facilitating the funding of large-scale projects, since those mechanisms are currently the primary means of funding such projects. There is a need for a mechanism through which input from innovators in research can be routinely collected and incorporated into institutional decision-making processes as well. A possible alternative to issuing PAs or RFAs for large-scale projects 3~ Personal communication, Richard Klausner, former NCI director.

OCR for page 80
FUNDING FOR LARGE-SCALE SCIENCE 129 aimed at particular topics would be to develop a special category, with specific review criteria and oversight requirements, for large-scale projects in general. Doing so would greatly speed the process for researchers with novel ideas while still maintaining a rigorous vetting process. A third possibility would be to make greater use of Defense Ad- vanced Research Projects Agency (DARPA)-type strategies for funding large-scale, technology-driven projects, as described in Chapter 3.NCI's Cancer Genome Anatomy Project and Unconventional Innovations Pro- gram could prove instructive in this regard. In any case, standardizing the methods for institutional oversight of such projects with regard to management structure and progress assessment over time would also improve the process, as is discussed in greater detail in Chapter 5. A fourth potential mechanism to speed and facilitate the launch of large-scale projects would be to set up a loan program through NIH for the purpose of developing scientific infrastructure, such as new buildings or the purchase of expensive new technologies for research. Such a pro- gram would allow extramural institutions to react quickly to changing needs and opportunities in the field by securing funds from NIH early on, and then repaying the loan through traditional fundraising activities. As noted in Chapter 3, several novel NIH programs have been launched in recent years in order to undertake large-scale research projects. These efforts depended on the institutional leadership at the time. Since many of those individuals have now left NIH, the future of such programs and the potential for launching other new programs is unclear. One way to reduce this variability is through long-term, Institute-wide strategic planning by the NIH director, as Elias Zerhouni is currently striving to do (Metheny, 2002; Kaiser, 2002~. This planning process incorporates input from Institute and Center directors, as well as from leaders among intramural and extra- mural scientists in both academia and industry. Such an approach provides the best opportunity to ensure that NIHis responding effectively to chang- ing needs in the field by funding innovative and useful projects in a timely fashion.