1. Introduction
Literature on industrial districts as well as other forms of spatial agglomeration of firms
(clusters, milieux innovateur, etc.) has stressed the systemic dimension of such productive
organizations1. In the case of industrial districts two systemic ‘levels’ have drawn the
attention of scholars: i) the inter-firm links based on customer-supplier relationships along
a production chain; ii) the ‘local system’, which is composed not only of firms but also of
social and political institutions. From the original contribution by MARSHALL (1920)
literature has identified several advantages associated with the spatial agglomeration of
firms: better access to specialized resources (information, skilled workers, suppliers);
complementarities among firms; better ability to generate knowledge (‘learning by
interacting’) and to innovate; access to specialized institutions and public goods; higher
efficiency due to local rivalry and peer pressure (PORTER, 1998; BOARI et al., 2003).
These aspects have been widely analysed both at a theoretical and empirical level.
Theoretical models of industrial districts are mainly qualitative. In these models a
substantial homogeneity of firms belonging to the district is normally hypothesized, or
taken for granted (BECATTINI, 1989). This assumption is also made in quantitative
models of spatial agglomeration (BOSCHMA and LAMBOOY, 2002; DURANTON and
PUGA, 2003). At the moment we lack theoretical treatment of the relationship between
firms’ characteristics and the structure and evolution of business clusters. Little is known
about whether the mechanisms at the basis of agglomeration economies influence firms’
characteristics and how the heterogeneity of such firms affects the structure and internal
organization of business clusters and their evolution over time.
Recent empirical investigations on Italian industrial districts have stressed the increasing
heterogeneity of district firms. They agree on a general tendency towards the concentration
of district output in the hands of a few firms (COSSENTINO et al., 1996; LAZERSON
and LORENZONI, 1999). At the same time the largest firms tend to achieve a better
control of the production chain tightening the relationships with their suppliers (DEI
OTTATI, 1996a; CORO and GRANDINETTI, 2001).
One of the main and quantifiable effects of these changes should be the increase in the
size of some leading firms belonging to industrial districts. The empirical evidence about
this is not straightforward. This is because the organizational form normally adopted by