Fernanda MAZZOTTA | Projects

FIRM ENTRY AND EXIT IN ITALIAN PROVINCES: INTERNATIONALISATION AND AGGLOMERATION ECONOMIES.

We carry out estimates of firm hazard rates of Italian manufacturing firms considering both the role of local agglomeration economies (Caianelli, Montresor, Marzetti, 2010; Fritsch and Schindele, 2010) and the role of global links related to exports, inward foreign investment and domestic firms investment abroad (Wagner, 2012). We use AIDA data (2002-2010), an extensive collection of information at firm level, and merge them with data from ISTAT (census and yearly series of “Provincial Accounts”). Thus, our analysis combines firms, industry and location specific determinants. The localisation variables are related to local agglomeration, entropy, specialisation (such as Theil and Balassa index), local industrial clusters. The span of time considered allows us also checking for the effects of the 2008 world crisis. The empirical application of our research refers to the Italian economy, where the geographical distribution of firms is historically characterised by a social-economic dualism between North-Centre and South. Italy is also one of the most popular country for agglomeration economies and industrial districts. The spatial agglomeration of firms is an important source of both positive and negative externalities, whose effect on regional economic performances depends also on their impact on the firms’ survival chances. The timing for testing the impact of local economies is appropriate as recent empirical studies on the Italian industrial districts have questioned that district firms show the superior resilience they had in the past (Bugamelli et al., 2009; CENSIS, 2010). On the other hand, the inclusion of variables related to innovation and internationalization let us identify the relationships between firms’, industry and local (province) characteristics and the firms’ international competitiveness, which recent studies have strongly confirmed to be crucially related to their demographic dynamics and growth (Altomonte et al., 2012).Our innovative dataset is being built by matching balance sheet data provided by AIDA with administrative data by Istat and Movimprese. It includes observations for approximately 100,000 firm observations per year.Thus, we estimate a hazard model where the probability of firm exit is modeled as a function of different determinant of agglomeration economies (population density, industry specialization, districtuality index, employment in districts, intra-industry variety, unrelated variety, related variety indices) and three different types of firm according to internationalization of the property: national firms, foreign multinationals, domestic multinationals, controlling for Pavitt clusters, sector technology classes, sector export intensity, in addition to other standard variables in duration models at firm (size, productivity, capital intensity, labour cost), industry (market barriers to entry such as minimum efficient scale, concentration, mark up) and province level (death rate, start up rates, patent rates, wage costs, labour productivity, financial risks, unemployment rates). Next we intend to deepen the analysis of the following most important issues: 1) Validation of our dataset with ISTAT data & further cleaning; 2) merging with other dataset from ISTAT providing information on subcontracting, exporting, delocalisation, inter firms linkages at global level; 3) testing for other classifications and for different features of industrial districts; 4) deeper investigation of the link between firms local bonds(multinational and ID) and global activities and firm exit over the crisis ; 5) check for input-output relations between firms or industries.