R&D,patenting and innovative activities: A statistical exploration

Better understanding of, and policies towards, technical change, requires better measurement of technical change. No single measure is perfect. Taken together, statistics on R&D and on patenting activities give important clues about the rate and direction of innovative activities, and also show the dangers of too hasty interpretation based on one measure. They both show a neavy concentration of innovative activities in chemicals and engineering (electrical, electronic and non-electrical) sectors; relatively rapid growth of innovative activities in drugs, scientific instruments and food products, and slow growth in aerospace and electrical products, in the USA between 1963 and 1974. They both show relatively high levels of innovative activities in Germany, Sweden and Switzerland, and relatively rapid rates of increase in Japan and Sweden between 1967 and 1975. They also show the strong association in the chemicals and engineering industries between the levels of innovative activities and of export competitiveness.On the other hand, taken together they suggest that patent statistics underestimate innovative activities in large firms, and that R&D statistics do so in small firms. This casts doubt on widely held assumptions about diminishing rates of innovative activity in very large firms, and about the non-electrical machinery and fabricated metal products sectors as “traditional” and “non-innovative” industries. Four factors are put forward to explain differences in what is shown by the patent and by the R&D measures: First, competitive behaviour, with smaller firms making a relatively greater use of patents, and bigger firms of R&D activities; second, different degress of specialisation and formalisation of innovative activities in and around R&D departments; third, variation across sectors in the degree to which patents measure an increment of technical improvement; fourth, institutional factors in aerospace and other defence-related sectors, and in motor vehicles, where patenting is low and the proportion of routine testing in R&D comparatively high.