Research and Product Innovation in Australian Manufacturing Industries

Abstract

EMPIRICAL studies of research and development (R & D) activity in manufacturing typically proceed on the assumption (often implicit) that measures of R & D inputs are satisfactory proxies for the output which results from industrial research. For example, most investigations of the relationship between a firm's innovativeness and either its size or market power employ expenditure on R & D, or R & D employment as surrogates for innovativeness, despite the obvious point that they are measures of resources devoted to R & D activity in the firm and not of the results of that activity.' The explanation for this common procedure is straightforward. R & D inputs are easy to measure, and statistics of R & D employment, R & D professional employment and R & D expenditure are now widely collected. By contrast, the very concept of R & D output is both complex and elusive, and hence measurement poses many difficulties. The few existing studies of the relationship between R & D inputs and R & D output have used either patent statistics or the sales of new products as measures of R & D output. In a study of American patent data and R & D inputs across several industries in the early I960s, Mueller concluded that 'there appears to be a significant relationship between what goes into the inventive process and what comes out of it' [4, p. 36]. Several studies of technological change have been made of the U.S. pharmaceutical industry for which there exist data on the proportion of sales attributable to new products. Comanor [I] and Vernon and Gusen [7] employ these data in inquiries oriented more to an examination of the R & D :firm size relationship than to an explicit consideration of some aspect of the underlying research production function. Comanor and Scherer [2], however, compute the direct correlation coefficients between alternative measures of research activity in the U.S. pharmaceutical industry, including patent and new product sales as well as R & D inputs. Together, these studies have led to the