Generic Policx Issues 13 



(Index; 1967=100) 



1960 62 64 66 



70 72 74 76 77 



FIGURE 6. Relative Change in Productivity' in Manufacturing Indus 

 tries by Selected Countries; 1960-77. 



'Output per worker hour 



Note: Estimates are shown for latest year 



Source; National Science Foundation. Science huliauors. 1978. 



ratios of R&D spending to sales (for example, chemicals, 

 electrical machinery) experience substantially higher 

 growths of productivity and output than industries with 

 low ratios (for example, textiles). Research results re- 

 ported in Science Indicators — 1980 confirm these implicit 

 correlations and trends. A study by Mansfield suggests 

 that there is a strong and direct relationship between the 

 amount of long-term basic research carried out by a firm 

 and its rate of productivity growth, even when the amount 

 invested in applied R&D is held constant.' 



Since the persistence of the relationship between ad- 

 vances in knowledge, innovation, and economic growth is 

 not easily documented, those estimates are subject to 



some uncertainty.^ For example, productivity is very diffi- 

 cult to quantify; approaches to its measurement depend on 

 interpretations of its meaning, which vary considerably. 

 In a purely quantitative sense, productivity can be viewed 

 as the ratio of outputs, in terms of physical quantities of 

 the product produced, to inputs, such as labor and capital. 

 But, those kinds of estimates do not take into account such 

 things as the quality of the product produced. Therefore, 

 quantitative indices of productivity that are based on the 

 number of items produced and do not show increases in 

 quality over time can give the impression of stagnation in 

 productivity advances. Indeed, the contributions of ad- 

 vances in technology to the qualitative aspects of produc- 

 tivity could be far greater than purely quantitative meas- 

 ures would suggest. Moreover, the problem of stimulating 

 industrial innovation cannot be divorced entirely from the 

 problem of the social utility of a particular innovation. 

 Despite these caveats, there is wide agreement that 

 there has been and continues to be a strong relationship 

 between research and development activities and eco- 

 nomic growth. Therefore, a major question that will need 

 continuing attention during the next 5 years is: How can 

 the United States maximize the contributions of science 

 and technology to the national goals of increasing innova- 

 tion, productivity, and economic growth? 



THE CAUSES OF LAGGING INNOVATION 



Stimulating innovation has been and continues to be a 

 major national goal, implying that current levels are be- 

 low those expected or desirable. In fact, there are indica- 

 tions that innovation and productivity in some American 

 industries are lagging behind those of several other indus- 

 trialized countries, such as West Germany. France and 

 Japan, and that the United States is thereby losing its 

 international preeminence in such vital industries as con- 

 sumer electronics, metallurgy, and automotives 

 (NRC-14; AAAS-2: TRANS). In the military sphere, the 

 Soviet Union appears to be closing the gap in implement- 

 ing such key technologies as electro-optical sensors, guid- 

 ance and navigation, hydroacoustics, optics, and propul- 

 sion (NS). 



Understanding the causes of the innovation lag in the 

 United States requires some dissection of the total indus- 

 trial innovation process. That process consists of a series 

 of stages ranging from the generation of new ideas, 

 through the development of pilot projects to further refine 

 and develop new technologies, to their conversion into 

 marketable products and processes. There does not appear 

 to be any decline in the ability of industrial scientists and 

 engineers to come up with innovative ideas. Rather, most 

 observers believe that the problem is inherent in the subse- 

 quent stages of the total innovation sequence — namely in 

 providing sufficient support to the R&D activities needed 



