Review of Industrial Organization 23: 301–314, 2003.
© 2004 Kluwer Academic Publishers. Printed in the Netherlands.
Adoption of Product and Process Innovations in
Differentiated Markets: The Impact of Competition
Chemnitz University of Technology, Department of Economics, D–09107 Chemnitz, Germany
Abstract. The paper examines the effects of the degree of competition on ﬁrms’ decisions to innov-
ate in differentiated markets. Firms favor product innovations if they produce close substitutes (so
competition is severe) and favor process innovations if products are differentiated (so competition is
less severe). Assumptions on the strategic complementarity of product and process innovations and
on the decreasing returns of a product innovation are found to be the critical assumptions in the sense
of Milgrom and Roberts (1994).
Keywords: Competition, process and product innovation, product differentiation
Every year, producers of basic fashion items have to decide whether they want to
introduce a new collection, to reorganize the production process for the present
collection, or to sell the present one with few changes for the upcoming season.
Whether the ﬁrm pursues a product innovation (new collection), undertakes a
process innovation (reorganization of production), or does nothing, has important
implications for the ﬁrms’ competitive position. It may be suspected that ﬁrms’
choice depends on the market structure, i.e., whether or not competition is intense.
On the one hand, introducing a new, differentiated product may allow ﬁrms to
escape ﬁerce competition. On the other hand, reducing costs for existing products
may be an aggressive strategy inducing competitors to exit the market altogether.
There are only a few theoretical papers studying the effect of product differentiation
on the choice between process and product innovations, and they provide divergent
results. This paper addresses the issue in a general framework that nests a broad
class of models. I will identify critical assumptions for the results in the sense
of Milgrom and Roberts (1994). The diverging results of previous papers may be
traced to alternative critical assumptions they make.