Treating patent as relational data: Knowledge transfers and spillovers across Italian provinces

The paper adopts a new perspective to investigate the characteristics of innovation clusters within and across Italian NUTS3 regions (province). Patents are used as relational data connecting inventors to applicants along a dual interpretation of a “knowledge production” and a “knowledge utilization” function. The paper intends to map and measure the structure and the evolution of a series of innovation sub-systems (both at territorial level and at the industry level) in two different periods of time, 1987-1991 and 1997-2001 to detect how networks of inventors and applicants have changed over time. The paper uses the CRENoS database on regional patenting, built on EPO data, spanning from 1978 to 2001 and uses two complementary analytical tools (network analysis and spatial econometrics) in order to analyse the behaviours of inventors and applicants within 103 Italian NUTS3 regions (province) and 5 specific industries chosen according to the Pavitt’s taxonomy (Footwear, Textiles, Chemicals, Personal Computers and Machine). The paper deals with the following research questions. Is the polarisation of the geographical structure of the innovative activity in Italy, replicated both at the inventors and at the applicant level? Which are the relevant changes in the 20 years? Can we distinguish between a “knowledge production function” and a “knowledge utilisation function”? Which is the structure of knowledge flows between inventors and applicants in Italian provinces? Are there industry-specific differences? Which are the determinants of the knowledge flows between Italian provinces?