Academic research and industrial innovation: An update of empirical findings

TLDR

Earlier studies estimated that a significant share of new products and processes in sectors such as pharmaceuticals, information technology, chemicals, and metals stem from academic research conducted within 15 years of commercialization. This paper extends those estimates by collecting data from 77 major firms for 1986–1994, replacing the oil sector with machinery. The authors surveyed these firms to quantify the proportion of innovations derived from recent academic research. The 1986–1994 results mirror those of 1975–1985 overall, though industry‑specific variations are evident. Industry means are reported in Table 1.

Abstract

In 1991, I published in this journal estimates of the percentage of new products and processes based on recent academic research—that is, academic research occurring within 15 years of the commercialization of whatever innovation is being considered— in the drug and medical product, information processing, chemical, electrical, instruments, metals, and Ž . oil industries in 1975–1985 Mansfield, 1991, 1992 . Because there has been considerable interest in such estimates , I have gathered similar data from a sample of 77 major firms for 1986–1994, the only significant difference in the sample design being that the machinery industry, which is reasonably active in this regard, has been substituted for oil. 3 Table 1 shows that the overall results for 1986– 1994 are very similar to those for 1975–1985, although particular industries sometimes vary substantially, as one might expect. The industry means in