GEOGRAPHIC CONCENTRATION AND VERTICAL DISINTEGRATION IN SERVICES. EVIDENCE FROM THE METROPOLITAN AREA OF MILAN

A recent strand of the economic literature has emphasised the role of services, and in particular knowledge-intensive business services (KIBS), as a primary source of knowledge creation and diffusion. Since this transferring process often occurs through strong supplier-user, face-to-face, interactions, the role of the local concentration of industry, and of spatial proximity in particular, becomes crucial. Theoretical and empirical literature show that the geographic concentration of industry induces firms to vertically disintegrate their production, due to the lowering of transport and governance costs as well as to the reduction of opportunism in managing transactions (Love and Roper, 2001; Helsley and Strange, 2007). However, the evidence is primarily based on manufacturing firms, whereas little or no attention is given to service firms. In this paper we try to fill this gap by assessing the effects of different measures of spatial concentration on service firms’ vertical disintegration within the metropolitan area of Milan. Data used in this work are drawn from AIDA, a commercial database collected by Bureau Van Dijck and gathering information on balance sheets data as well as longitude and latitude of Italian joint stock companies. Relying on this rich firm-level dataset, we build a sample of more than 54.000 service firms located in the Local Labour System (LLS) of Milan in year 2008. In the spirit of Arzaghi and Henderson (2008), we first geo-referenciate our data by employing a GIS routine. In this way, we identify the exact geographical position of each service firm in the LLS. Then, we define a set of rings moving out of increments of 250 metres, and we count the number of firms located within each ring. In such counting, following Holmes (1999) and Li and Lu (2009), we define two variables measuring geographic concentration of service industry. For each firm, we compute, ring by ring, the number of neighbouring firms that are in the same two-digit industry, and the number of firms that are in all the two-digit industries except for the one in which the firm operates. In this way, we estimate the impact of proximity-based specialization Vs diversification economies on service firms’ vertical disintegration, exploiting information on the actual distance between each pair of firms in the sample. As additional regressors, we also include firm size, age and their squared terms. Our dependent variable, instead, is calculated as the share of business services actually purchased by firms over total production costs. This purchased-inputs variable allows accounting for the fact that “many business services are likely to be exactly the kind of locally produced intermediate input that producers in localized areas will have greater access to than producers in isolated areas” (Holmes 1999, p. 316). Preliminary results confirm that service firms tend to vertically disintegrate more if they are located within a dense urban area, this being particularly true for KIBS. In line with previous literature on manufacturing, we still find that geographic agglomeration externalities tend to decay relatively fast in space.