50 



/ 1 ll'Rf > 1 7 \ 6 THL' A LLOCA TIOS PROCESS / 25 



Within the General Constraints Determined by National Priorities, 



the Selection of Individual Projects Must Reflect the 



Standards of the Scientific and Technical Community. 



(Recommendations 10 and 11) 



RECOMMENDATION 10. Because competition for funding is 

 vital to maintain the high quality of FS&T programs, competitive 

 merit review, especially that involving external reviewers, 

 should be the preferred way to make awards. 



The h!j»hest-c|ualir\ projects and people should be supported with FS&T 

 funds. The best-known mechanism to accomplish that is some form of open compe- 

 tition involving evaluation of merit by peers. Competitive merit review involves the 

 use oi criteria that include technical qualir\, the qualifications of the proposer, 

 rek-^ance and educational impacts of the proposed project, and other factors per- 

 taininji to res<rarch goals rather than to political or other nonresearch consider- 

 ations."'" Open competition means that, at some level within the framework of an 

 aj»cnc\"s mission, researchers propose their best ideas and anyone may apply and be 

 funded regardless of institution or geographic location. However, in the ca.se of 

 highly targeted missions. qualit> can also be maintained by knowledgeable program 

 managers who have established external scientific and technical advisory groups to 

 help assess qualitx and to help monitor whether agency needs are met (see Supple- 

 nunt .S and Box 11.8). 



Tile c«)mmittce believes that the principle of merit review — which empha- 

 sizes competition among ideas, diversity of funders and performers of research and 

 deveiop'.iicnt.and organizational flexibilin — has been largely responsible for the 

 remarkable quality, productivin. and originality of U.,S. .science and technology in 

 the p.isi. Competitive merit review should be the method of choice for making 

 future decisions about RS&T funding. 



.Many federal research and development agencies have developed some form 

 ol competitive merit re\ iew process to use in making extramural awards for re- 

 searcii. trainmg. and facilities. The\ have also worked to develop equivalent systems 

 of review for allocating intramural funding, but merit review of in-house research is 

 n;iich mor.. difficult because federal research scientists and engineers are in the civil 

 ser\ ice and still retain salary and benefits even if they are not productive or their 

 area has lower prioritv or has become obsolete. That problem is a perennial one in 

 the pc-iodic reviews of federal laboratories.*" The FFRDCs, including the national 

 lahoratorifs. also have procedures for allocating research funding competitively 

 based on performance. .Some do it well, but overall the results have been uneven."' 



There are other approaches to promoting high quality' in federally supported 

 research and development. Some programs try to identif\' top researchers and give 

 tiiem long-term support rather than require them to submit specific proposals to 

 compete evcr\ few years Some funding for agricultural research is allocated to 

 state agricultural experiment stations and land-grant colleges on a formula basis, and 



