عوامل اثر سرریز دانش فضایی در استان های ایتالیایی

Statistical evidence suggests that the relevance of knowledge spillovers has increased over time. In this paper we focus on regional knowledge spillovers and adopt a new econometric transformation that allows inference on potential inter-regional knowledge spillovers, accounting for spatial interdependencies. Determinants of inter-regional knowledge spillovers are explained with a sample of 103 Italian provinces. We find that a region’s absorptive capacity, measured by local R&D expenditure and social capital, implies a reduction of outward knowledge spillovers. Identification is based on the use of Two Stages Least Squares and Fixed Effects estimates.

In the economics literature, the last 20 years have been marked by countless studies on the notion of knowledge spillovers (henceforth, KS). KS can be defined as a non-rival knowledge market externality, by which the positive effects of the invention of a new technology, product or process spill over to other organizations. This holds true both for firms as well as for spatial units (cities, provinces, regions or countries). Operationally, the identification of KS has been difficult. In the 1990s two major empirical advancements have been made in this direction. On the one hand, improved spatial econometrics techniques have allowed researchers to estimate spatially lagged parameters of R&D investments, knowledge, and human capital, which can all be interpreted as potential KS (e.g. [1]). On the other hand, the intuition provided by [2] that knowledge may leave a “paper track” when moving around in the form of citations gave rise to a burgeoning literature seeking to identify KS by means of flows of citations.

Understanding how new knowledge is created and adopted, and tracking its diffusion in time and space is a crucial issue for knowledge-based economies. In this paper, we used a novel spatial econometric technique to gauge the magnitude and determinants of outward knowledge spillovers, while employing a newly constructed data set on Italian NUTS3 regions. By accounting for the specific peculiarities of the Italian case (i.e. the presence of an irregular labor force), and using micro-data on social capital components, along with traditional economic variables, we thus provide a detailed view of knowledge and productivity spillovers’ dynamics.