Productivity patterns in the Italian manufacturing sector

Even if it is recognized to be a key factor for the purpose of explaining regional development patterns, most of studies on economic convergence omit to consider the crucial role played by the industry mix. The recent empirical literature, analyzing the Italian experience, reveals that regional differences in productivity are the main reason for regional inequality in per capita income in Italy in analyses focused on regional data at an aggregate level. Moreover, quite different productivity patterns emerge when the analysis is conducted by considering sectors and geographical areas (Southern and Center-Northern regions) separately (Paci-Pigliaru, 1995; Di Liberto et al., 2006). Heterogeneous productivity levels may come from differences in sectoral composition or may be uniformly distributed across sectors. This paper aims at investigating the role of differences in the sectoral composition of activities that are associated to the regional specialization, and in productivity gaps that are uniform across sectors, in explaining inequalities in aggregate productivity per worker across Italian regions, with a focus on the manufacturing sector. In addition, we identify the potential determinants of the differential in regional value added per worker and analyze the role of each source of regional differential productivity within a growth theory framework. We proceed by formalizing an extended conditional convergence equation to explicitly account for a decomposition of per worker value added in manufacturing which is derived by means of a modified shift-share approach following the work of Esteban (2000). We then proceed by assessing the role of these determinants in a growth framework by introducing a SUR analysis which treats the relationship that describes the evolution of all components as a system of seemingly unrelated regression equations. Our analysis is concentrated on the evolution of the Italian economic development process in the manufacturing sector because: (i) its pattern in Southern regions seems to show a clear tendency to diverge from the national average; (ii) it is a typical footloose sector which is not characterized by a specific localization of productivity factors; and (iii) non geographical localization factors can be relevant. Finally, the Italian manufacturing sector is one of the most analyzed and is characterized by a well-known remarkable degree of regional heterogeneity in variables, such as per worker income levels and human capital stocks. The analysis is carried out by using a dataset containing information at regional level for seven sectors of the Italian manufacturing industry covering a thirty year period (1970-2000)