ABSTRACT. The purpose of the present paper is twofold.
First, we formulate a general method of identifying clusters
of economic activity in a country or region. The methodology
is based on generally available industry data and does not start
from a priori notions of geographic or technology clusters.
Second, the proposed methodology is then used to identify the
main clusters of economic activity in Ohio and Sweden and
their evolution over the last twenty years.
For several years now, a group of Swedish
researchers has investigated the role of “techno-
logical systems” in economic growth (see
Carlsson, 1995 and 1997). Technological systems
refer to networks of agents in any given field of
techno-industrial activity, interacting within a
particular institutional infrastructure, to generate,
diffuse, and utilize technology (Carlsson and
Stankiewicz, 1991). Thus, a technological system
is a form of cluster of economic activity, closely
related to Erik Dahmén’s concept of development
blocs (Dahmén 1950 and 1989), and similar in
many ways to the type of clusters studied by
Michael Porter in his
Competitive Advantage of
(1990). The technological systems frame-
work includes not only market interaction among
firms but also non-market interaction (especially
in the form of knowledge spillovers) among firms
on the one hand and various components of the
infrastructure (e.g., academic institutions, research
institutes, financial institutions, government
agencies, and industry associations) on the other.
A common feature in various approaches to
cluster analysis is the ad hoc nature of selecting
economic activity for study. But what if it is not
clear a priori what economic activities are clus-
tered together? There may be forward and
backward linkages, and formal as well as informal
networks, which may not be easily identifiable by
simply observing existing data, but whose nature
may be an important determinant of economic
development of a region. The input-output links
between, say, the steel industry and the down-
stream metalworking industries are well-known,
but the same is not true for the interaction between
manufacturing industries and service industries,
especially in the area of business services. The
relationships between new firms in emerging
sectors and existing (older and larger) firms in
traditional industries are of particular interest in a
The purpose of the present paper is twofold.
The first is to propose a systematic method of
identifying clusters of economic activity. The
basic idea is to identify industry clusters in a
country or region, based on generally available
industry data, through the use of a common
methodology which does not start from a priori
notions of geographic or technology clusters.
Second, we will apply the chosen methodology to
the analysis of two regions of approximately the
same size and industrial structure to see whether
the results are in line with expectations. Possible
avenues to further refinement of the proposed
methodology will also be elaborated.
Industry Clusters in Ohio
and Sweden, 1975–1995
Small Business Economics 12: 279–293, 1999.
1999 Kluwer Academic Publishers. Printed in the Netherlands.
Final version accepted on November 11, 1998
The Research Institute of Industrial Economics
114 85 Stockholm, Sweden
Case Western Reserve University
Weatherhead School of Management
Department of Economics
Cleveland, Ohio 44106-7206