The traditional analysis of innovation has focused on the Schumpeterian hyp
othesis of a positive link between market power and innovation. This often
includes an implicitly linear view of the innovation process, with R & D as
a necessary first step. This paper widens the determinants of innovation b
eyond R & D to include technology transfer and networking effects, thus ext
ending the standard Schumpeterian analysis. When tested on a dataset of c.
1300 UK manufacturing plants, R & D, technology transfer and networking are
found to be substitutes in the innovation process, with the two latter int
ensities especially important in increasing the extent of innovation. There
is no evidence that (actual) monopoly power increases the extent of innova
tion, but there are significant plant and sectoral effects on innovation.