Policy-makers working at the national and regional levels could find the territorial mapping of research productivity by field to be useful in informing both research and industrial policy. Research-based private companies could also use such mapping for efficient selection in localizing R&D activities and university research collaborations. In this work we apply a bibliometric methodology for ranking by research productivity: i) the fields of research for each territory (region and province); and ii) the territories for each scientific field. The analysis is based on the 2008-2012 scientific output indexed in the Web of Science, by all professors on staff at Italian universities. The population is over 36,000 professors, active in 192 fields and 9 disciplines.
The localization of universities within a particular nation has historic, economic, and sociological origins, and more recently is ever more influenced by policy and strategic decisions. Whatever the origin of the current territorial distribution of new knowledge suppliers, the policy maker certainly has interests in monitoring the evolution of the efficiency of research activities, for purposes of understanding and decision-making regarding the selective allocation of public resources. Similarly, for research-based companies in the private sector, the territorial mapping of research productivity by field can inform efficient choices in the localization of R&D activities, and research collaborations.
In the literature, the characterization of the scientific profile of a given territory is typically conducted by gathering and analyzing bibliometric data: specifically by analyzing the geographic distribution of scientific production, as indexed in the major bibliometric databases. This approach assumes that scientific publication in international journals is the principal form of dissemination of results from research activity, as conducted by universities and research institutions in general. Frenken et al. (2009) offer a particularly useful review of the full range of scientometric studies analyzing the spatial dimension of scientific production, beginning from the pioneering works by Narin and Carpenter (1975) and Frame et al. (1977). This latter work, under the suggestive title “The distribution of world science”, and based on data from the ISI Science Citation Index, maps the distribution of output from 117 countries and in 92 disciplines, over one year (1973). More recent studies, employing similar methodologies, have primarily concerned the spatial concentration of scientific production, which seems to have remained high for the industrialized nations of the OECD. These nations thus continue to account for the major share of world output (May, 1997;Adams, 1998;Cole and Phelan, 1999;Glänzel et al., 2002;King, 2004;Horta and Veloso, 2007), despite a rapid increase in scientific production from China (Leydesdorff and Zhou, 2005). Analyses at the regional level have been less frequent: one case is the work by Matthiessen and Winkel-Schwarz (1999), on the analysis of aggregated publication records for European metropolitan areas, for the years 1994-1996. Some scholars have also proposed analyses based on the spatial distribution of highly-cited publications, primarily for the identification of centers of excellence at the regional level (Bonitz et al., 1997;Batty, 2003). More recently, a work by Bornmann and Leydesdorff (2011), based on the Web of Science (WoS) data, identifies cities where top-10% highly-cited papers were published more frequently than would be expected, offering visualization of the results via Google Maps. In very similar manner, Bornmann et al. (2011) present methods for mapping centers of excellence around the world, in this case using Scopus data. Excellence in single scientific fields is identified, revealing agglomerations in regions and cities where highly-cited papers (top-1%) were published. Shifting the focus from cities to regions, Bornmann and Waltman (2011) use visualization methods (density maps) to detect regions of excellence at the global level, focusing on the top 1% of 2007 papers indexed in Scopus. Very recently, Bornmann et al. (2014) presented a web application to identify research centers of excellence by field worldwide, using publication and citation data.
Within Italy, Tuzi (2005) pioneered bibliometric measures of the scientific specialization of regions, by two separate indicators: one based on publications and the other on average citations per paper. Morettini et al. (2013) document “knowledge activities” at the regional level, through the measurement of R&D expenditures, patents, and publications originating from “local labor systems”. Abramo et al. (2009) have mapped the centers of excellence in Italy by analyzing the concentration of top scientists in the same institution. The same authors have recently presented a bibliometric methodology to carry out a spatial analysis of the impact produced by research institutions (Abramo et al., 2015), and to identify the scientific specialization of territories (Abramo et al., 2014b). The first contribution ranks territories in each research field by the total impact produced by local institutions, while the second one measures the “scientific” comparative advantages of territories. Both are held to have broader significance, most notably for the methodological approach. The authors use field-normalized citations, and not simply the counting of publications, to map the territorial distribution of new knowledge produced and the scientific specialization of regions: in fact, counts of publications alone do not permit an assessment of the real value of the new knowledge produced.
Continuing from their preceding works
This content is AI-processed based on open access ArXiv data.
