Research and Productivity in Wheal and Maize: An International Analysis 59 



to be underestimated, although the importance of the omitted in- 

 puts should not be exaggerated. An agricultural production pro- 

 cess can be viewed as a mixture of two processes: biological and 

 mechanical, with a production function 



G=/%(.x,)./„,(.v„)) 



where Q is output; b the subscript of the biological process; m the 

 subscript of the technical, mechanical process; x,, x vectors of 



b m 



mput. 



The inputs into the biological process (seeds, fertilizers, water, 

 correct choice of methods, and timing) determine the potential 

 yield. Labor and machinery are the inputs into the mechanical 

 process; factor substitution is possible within the mechanical pro- 

 cess, although there is very little substitution between the two 

 processes.^ The research measure is limited to biological work. 

 By omitting labor and capital the analysis is restricted to the 

 biological subprocess. 



The omission of fertilizers, water, and perhaps also an index of 

 seed quality, pesticides, and similar biological inputs is more 

 serious. Their omission is justified by the extent to which the 

 adoption of these inputs is due to agricultural research. We are in- 

 terested in the total effect of research, including the indirect con- 

 tribution through other inputs. However, these omissions will 

 bias the estimates of research contribution upward. 



An international comparison offers the opportunity to include 

 "borrowing" — the transfer of knowledge — into the analysis. It is 

 important to recognize the differences in the transferability of the 

 major aspects of knowledge in wheat and maize. Since wheat is 

 self-pollinating, regularly harvested grains can be used as seeds 

 and new varieties will be easily "diffused" from farmer to farmer. 

 Hybrid maize seeds, on the other hand, have to be propagated by 

 specialized agencies and distributed to the farmers. Moreover, 

 wheat is much less locality-specific than maize. Mexican wheat 



3. The hypothesis of the separation of the agricultural production function into 

 two processes cannot be tested here (Sandan 1970). This hypothesis is consistent 

 with the recent analysis of Hayami and Ruttan (1971), whose major finding can be 

 interpreted to mean that, because of the corresponding factor-price relationships, 

 American research was directed to advancements in the mechanical process 

 while the Japanese was geared to the biological one. 



