Are there Schumpeterian waves of innovations?

This paper responds to a recent criticism of the empirical evidence of bunching of innovations. An examination of various long-run innovation samples shows that there is indeed very poor evidence of innovation waves in the time before the mid-19th century. Thereafter, however, two long waves of major innovations occur, both having a lead of approximately 10-15 years over the economic long wave as identified in an earlier study. A t-test confirms that these waves can be clearly distinguished from random fluctuations. In the final section some suggestions for further research are outlined.