THE LESSON LEARNT FROM Klevorick et al., 1995. On the sources and significance of interindustry differences in technological opportunities

In this paper, the authors argued why the R&D intensity is high in some industries and low in others. In previous researches, it was demonstrated that (1) firm size and market structure or (2) market size and growth in demand is the reason. But here, they argued that the technological opportunity and the ability to appropriate returns from new developments are crucial reasons.

By technological opportunities, they comprise the set of possibilities for technological advance which could be measured in terms of the distribution of values of improved production-function or production-attribute parameters, or of the distribution of returns to R&D (p. 188). These opportunities illustrated that the R&D at the level of the firm or industry is subject to diminishing returns. "As resources are devoted to R&D and projects are completed, technological opportunities are depleted and the pool of opportunities can be exhausted" (p. 188). Thus, an industry which was thought rich in technological opportunities was the one where high R&D intensities and high rates of technical advance tend to be sustained over time.

Basing on this logic, the authors thought there were 3 types of sources of technological opportunities: (1) the advance of scientific understanding; (2) technological advances originating outside the industry; and (3) feedbacks from technology.

First, it was believed that significant technological breakthroughs can be traced directly to advances in basic general scientific understanding that occurred just prior to the breakthrough (p. 189). After facing the need or an object, the science will be employed from the mind of the researcher. Thus, scientific advantage generated offsets to diminishing returns to R&D. Meanwhile, the scientific and engineering disciplines could provide and help facilitate technical progress of various kinds. These basic researches could be applied in various industries.

Second, here the party at the upstream or downstream of the supply chain could be important to industry R&D efforts. Third, the technology and research of today might only provide new starting points for future technology, which was called "compulsive sequences" or "natural trajectories." The logic of this feedbacks also suggested that although they could partially offset the tendency of R&D to deplete prevailing opportunities, in the long run they cannot totally offset that tendency (p. 192).

By using the data from Yale Survey, the above proposition was verified. Among the industries, the industry like drugs and semiconductors appear close to science; while motors, generators, industry controls, motor vehicle parts, and accessories are relatively distant from science. Universities and labs are often important to new sciences, but the basic sciences, compared with applied science and engineering fields, did not show high relevance to industry R&D (p. 197).

Material suppliers, production equipment suppliers, research equipment suppliers, and users were found important sources of technological knowledge. Namely, the sources with vertical linkages were important. About the natural trajectories, changes in scale of production, mechanization/automation, improving process yield, improving input material, and from batch to continuous process are found important resources.

Accordingly, the authors could distinguish industries that are rich in technological opportunities from those that are not (p. 201). Also among high opportunity sectors, different sources of opportunities impacted differently on R&D/sales, process innovation, and product innovation. The R&D intensity was correlated with (1) the connections with several of the fields of science; (2) the contributions made by university and government labs; (3) the work of equipment suppliers.