Several things come out immediately - none of which are counterintuitive. 

 First, given present technology, there is a point at which further increments 

 in the budget do not provide significant increases in the amount of new informa- 

 tion. Second, it is difficult to increase the value of real time information 

 (0 to 2-year horizon) without an inordinate expenditure of funds. Third, the 

 time horizon from 3 to 7 years is an area within which it is difficult to 

 modify the programs to achieve significant specific increases in information. 

 Fourth, the distant horizon tends to accrue benefits from programs in the 

 middle range and of itself does not need to be addressed in explicit terms. 

 This is probably just as well since it is at this horizon that the environmental 

 factors begin to dominate and it is here also that there is little capability 

 at present to effectively predict what will happen 10 years from now anyway 

 insofar as environmental change is concerned. 



Going beyond this level of generalization would not be particularly helpful. 

 One can play other games with this system and get slightly different results. 

 Almost without exception, when these games are played, one discovers very little 

 that isn't obvious in terms of just plain common sense. Nonetheless, from the 

 standpoint of the administrator, whether the administrator is the Center 

 Director, the Director of a Service, or the Director of NOAA, one can now come 

 back to the "how much is enough" question. 



Let's go back to figure 1 and the performance function for fishing based 

 on survey cruises. The administrator can make the judgment that from his point 

 of view the value of the information being provided is not 0.46 but 0.6. He, 

 after all, is the one who makes decisions and his judgment of the value of 

 information is important and in the long run should probably be one of the 

 driving forces in program development. This has the effect of transforming the 

 X axis. The performance function itself hasn't changed, but the expenditure to 

 achieve that level of information is now addressed on a different basis. 

 Information with the value of 0.6 on fishing costs $6 million as far as he is 

 concerned. This is his "comfort index." Now one can have a constructive 

 dialogue. 



This performance function model is but an extension of our Northeast 

 Fisheries Center model for the ecosystem. For all practical purposes, it is 

 the same model. It has simply been modified to serve another purpose. Much 

 of the information to be presented at this symposium will deal with models at 

 the other end of the spectrum. These activities are important. They are, 

 however, critical to the development of these other versions. The version I 

 have presented here deals with broader problems, as for example, budgeting, 

 program development, and evaluating areas both for reduction or enhancement. 

 Such integrative and semantically different models have the virtue of allowing 

 the research administrator to get into the system and evaluate program needs 

 in terms of his perception of our reason for being, on a communication bridge 

 where both the expert and the administrator can meet as peers. 



33 



