This paper proposes that the specific pattern of innovative activities in a
n industry can be explained as the outcome of different technological (lear
ning) regimes. A technological regime is defined by the particular combinat
ion of technological opportunities, appropriability of innovations, cumulat
iveness of technical advances and properties of the knowledge base. Buildin
g upon the distinction between Schumpeter Mark I and Schumpeter Mark II ind
ustries, this paper provides empirical estimates of the relationships betwe
en indicators of the Schumpeterian patterns of innovation (concentration of
innovative activities, stability in the hierarchy of innovators and import
ance of new innovators) and indicators of the variables defining technologi
cal regimes.