Dynamic Analysis of sectorial Specialization and Agglomeration Economies in the Sicilian Local Labour Systems

This paper examines the spatial dependence among productive specialization, variety of the economic environment and knowledge spillovers. With the aim of identifying localization benefits in a particular area, some aspects of business settling are compared with trends in the development of economic activities and, particularly, with agglomeration process of the firms. On the basis of the model proposed by Ellison and Glaeser (1997) it is possible identifying an \u201cagglomerate space\u201d that is to say a \u201ccontinuous set of territorial areas specialized in a specific sector one of which, at least, proves to be provided with a level of specialization noticeably exceeding the expected value under the hypothesis of a lack of agglomeration benefits\u201d (Iuzzolino, 2004). The study of dynamic externalities will make a distinction between: 1) localization economies: intra-industry spillovers arising from the stock of knowledge transferred from one local firm to another belonging to the same industry. This type of externalities turn into productive specialization, skilled labour and enhancement of activities subsidiary to the core-business; 2) urbanization economies: inter-industry spillovers deriving from the localization of the firms in cities where the stock of knowledge was built up thanks to the historical diversity. The empirical analysis will be carried out pointing to the dynamics of employment consequent to the effects of the externalities above mentioned. Employment data are taken from the Manufacturing and Service Censuses of the years 1981, 1991 and 2001 and refer to the 77 Local Labour Systems (LLS hereafter) that represent the result of the spatial aggregation of neighbouring municipalities based on the daily commuting flows of local population owing to work reasons. Precisely, data pertain to the following sections of the classification ATECO 2002 (based on European NACE Rev.1.1): Mining and quarrying (C), Manufacturing (D) and Total private services (G, H, I, J, K). We will consider some suitable economic indicators to depict the characteristics of concentration: a) localization ratios for studying the evolutionary trends in specialization for different types of economic activity b) diversity measures, based on a Hirschman-Herfindahl index, suggesting the presence of Jacobs-type dynamic externalities or, on the contrary, the operativeness of MAR dynamic effects c) competition indices to gather information about the competition levels of the markets depending on their structure d) agglomeration indices to measure the geographic concentration of the economic activity consequent to high and/or increasing industrial concentration levels in the adjacent LLS