Abstract
There is a growing empirical literature investigating the relationship
between innovation activities of the firms and their ‘performance’. These
firm-level studies usually employ Heckmann’s methodology and go under the
name of CDM models (after the initials of the authors of the classical 1998
paper: Crepon, Duguet and Mairesse). The outcome of these studies is that
firms that innovate more usually perform better, but there is a considerable
variation when it comes to the details of the results. The aim of this paper is
to contribute to the methodology of such literature, firstly, by discussing the
theoretical foundations of the CDM models and thus shedding more light on
the hypotheses underlying them, and secondly by assessing the usefulness of
the Community Innovation Survey (CIS) in this respect. Several theoretical
approaches are invoked, including neoclassical economics, Porter’s framework
and evolutionary economics. It is argued that the positive relationship between
innovation and performance is strongly backed by economic theory but the
hypothetical mechanisms behind the link can vary. Investigating the nature
of the connection is a promising line of future research and examining the
parts of the CIS database that were largely neglected in CDM-type studies
to date can help resolve the puzzle.