Refining the federal-state partnership to address (a) the equity and efficiency of allocating federal funds for agricultural research, (b) priority setting and stakeholder involvement, and (c) public-private partnerships in agricultural research is constrained by the limited number of mechanisms used to provide federal support. An examination of alternative mechanisms may reveal new opportunities for the land grant system and its federal partner and complement the recommendations for change the committee has made with regard to specific elements of the agricultural research system.
At present, federal funding for extramural agricultural research is provided as formula funds, special grants for specific projects and institutions mandated by Congress, contracts, and competitive and peer-reviewed grants. Lesser-used mechanisms include cooperative agreements, small-business innovation research (SBIR) grants, and collaborative research and development agreements (CRADAs) with private industry. Although the variety of mechanisms suggests considerable flexibility, the system is, in practice, rather rigid for individual scientists or groups of scientists with a new, meritorious idea for research. This is true even for scientists who have experiment station appointments. A range of seldom-used alternative mechanisms does, however, exist, particularly in the design of competitive grants programs.
One way to reduce the paperwork and time consumption associated with processing competitive grant applications is to institute a preproposal step. Applicants would submit brief (2 or 3 page) preproposals for evaluation; the most meritorious would then be eligible for further consideration as full proposals. For example, if funding is available for 100 grants, only the most promising 200 to 300 preproposals would be eligible for resubmission as a full proposal. This would greatly decrease time spent on both proposal preparation and evaluation. Furthermore, this system can relatively quickly sort out proposals that do not fit subject matter criteria or are from unqualified investigators.
A related approach being tested in some NIH study sections is a system of triage. On receipt, standard competitive proposals are briefly evaluated by small committees, and the weakest one-third are rejected. This saves time because only two-thirds of the applications are fully reviewed, and the other one-third of applicants know of a nonfunding decision quickly rather than after the usual 6 months required for full evaluation. This permits rewriting and resubmission for the next round of funding, which is every 4 months at NIH.
Another mechanism used by some agencies, and under consideration by others, is the "chunk grant," in contrast to conventional full resource-recovery grants. Chunk grants might be for $50,000 per year for 2 or 3 years or might take the form of separate competitions for, for example, $20,000 or $50,000 portions of larger grants. With this scheme, the number of grants to be given is predetermined—hypothetically, 200 $100,000 grants; there is no negotiation over precise budgeting, budget justification, or overhead charges. Within the proposal there still might be considerable explanation of how funding would be used, but it is a much simpler approach and allows the investigator to allocate funds as deemed necessary.
Another appeal of this method is its honesty. Researchers cannot accurately predict what their precise financial needs will be 3 or 4 years after proposal preparation. Chunk grants and separate competitions eliminate the need for future-oriented guesstimates, and there still would be oversight and auditing to ensure funding is used legitimately.
Some perceive a considerable fringe benefit from such grants in that many investigators would be quite productive with $50,000 per year. This approach could ultimately