Size and Performance of
Local Clusters of Firms
Small Business Economics
17: 185–195, 2001.
2001 Kluwer Academic Publishers. Printed in the Netherlands.
ABSTRACT. The aim of this paper is to propose an
econometric application in order to detect the factors that
affect the size and the performance of local clusters of small
and medium firms (SMEs) such as industrial districts. Our
purpose is to extend some basic issues of models in economic
geography in order to account for further micro components
that seem fostering the rise and performances of local clusters.
We succeed in stating that, in addition to the home market
effect and transport costs, firms belonging to industrial
districts improve their competitiveness by exploiting some
assets, such those involved by ad-hoc services, generated
inside the district by the co-operation among firms. Moreover
looking for fixed effects in the study of firms’ competitive-
ness allows us to determine that sectorial differences do not
affect substantially the behavior and performance of SMEs.
Recently, one has witnessed a growing interest in
economic literature in joining the theory of
comparative advantages with that of the location
of firms. This new stream of literature (see
Audretsch, 1998; Fujita et al., 1999; Malberg et
al., 1997) has been principally developed for
looking at the conditions that allow the rise of
local agglomerations in regions or countries.
Nevertheless, the basic features of this new
approach may be fruitfully applied in tackling
some topics even for more micro agglomeration
structures as local clusters of firms.
This paper aims at developing an empirical
investigation on the determinants of export
flows for local clusters of firms. In particular, we
concentrate on a sample of local industrial dis-
tricts. For each of them we test an econometric
model in order to disentangle the original and
common features that affect the intensity of their
trade flow. Local districts we are referring to are
better known as Marshallian districts. Indeed they
are not uniquely a simple cluster of small and
medium enterprises. They display a high degree
of co-operation (and communication) among local
firms that entails the creation of local services
which district firms benefit from (Cooke and
Morgan, 1998). The geographical proximity, the
trust components (attached to a socio-cultural
identity) and the presence of local services allow
small and medium firms, associated to a network,
develop a sectorial specialization (joined with an
innovative capacity) that assists the competitive-
ness of the enterprises belonging to districts.
Since the spatial component takes an important
role, our econometric applications rely mainly on
some of the principal features embodied in the
gravity models, but, at the same time, we try to get
ahead of them extending the range of the field
these models concern. Indeed in gravity models
the intensity of trade flows results to be affected
by the size of the home market and transport costs.
Conversely, we aim at proving that in a more
micro setting there exists further local factors that
actively foster the export flows.
This micro contribution is concentrated on data
of Italian industrial districts which belong to a
single region (Lombardia). In this case, we focus
on the direct relationship between the attractive-
ness of the districts and their international
competitivity. As a consequence, we need to use
very detailed information at a district level and,
up to now, only Lombardia (which displays about
21 local districts in it) makes them available.
Comparing the size of Lombardia’s districts across
four years (with a classification by sectors of
activities), a pseudo-gravity model is applied. We
expect to confirm that the districts size directly
Final version accepted on March 22, 2001
IRES-Départements des Sciences Economiques
Université catholique de Louvain
Place Montesquieu, 3
1348 Louvain la Neuve