185 



11 



as a 'concern' for practical problems of clientele groups and applied 

 researchers and not focused solely on current fads, Journal articles, or 

 personal enhancement of reputation. 



They also show that the current shift of SAES units to competitive grant 

 and contract funding, as opposed to formula funding from federal sources and 

 regular state government appropriations, reduced the productivity of public 

 agricultural research, measured as the impact of research on state agricultural 

 productivity. It is possible that other socially valuable research outputs are 

 produced in large enough quantities to offset these lower local productivity 

 effects, but these other effects have not as yet been quantified. 



Furthermore, the Huffman and Just (1993) study concludes that one of the 

 most important long term agricultural researcl) managemeflt strategies is to 

 maintain a vertically integrated agricultural science establishment. In the 

 short run, pretechnology sciences and applied sciences may seem to be somewhat 

 independent. In the long term, technological progress can only occur when 

 vertical linkages across the layers of science and technology exist and 

 function well. Pretechnology scientists are more productive when they are 

 aware of practical needs, and applied scientists are more productive when they 

 are informed and literate In the parent sciences. Historically, the applied 

 sciences in agriculture have frequently become disconnected from the parent 

 sciences due to increased specialization of efforts and growing scientific 

 sophistication of general and pretechnology sciences. For example, 

 agricultural faculties of land-grant universities and USDA research agencies 

 were poorly prepared in the early 1980s to assimilate new scientific advances 

 In molecular and microbiology leading to the new biotechnology. 



In the pretechnology sciences, scientist-directed research programs are 

 more efficient than administrator-directed ones. Predicting advances In most 



