![Sample our Politics & International Relations journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5947/TOC-Subject-Sample-2021-Politics-International-Relations.jpg"})
ABSTRACT
This paper proposes a novel index of regional skill-relatedness and calculates this measure for all Norwegian labour-market regions. Studies of regional related diversification rely on measures of related variety, which build on the industry classification hierarchy. However, the growing literature identifying similarities in knowledge and competences across industries demonstrates that these classifications fail to identify a great deal of actual skill relatedness, and that measures based on empirical measures of industry relatedness are required. The skill relatedness measure builds on labour mobility flows across industries to develop a relatedness matrix for Norwegian industries. It further uses social network analysis to identify the number of other regional industries to which each industry in a particular region is related. Comparing this measure to the related variety index, the analysis shows that the two measures are highly correlated, but that the regional skill relatedness index is able to identify more of the relatedness across industries. In particular, the related variety index tends to underestimate the level of relatedness in many of Norway’s most technologically sophisticated manufacturing regions, whereas these rank highly in the regional skill relatedness index. Consequently, the regional skill relatedness index represents a promising new tool for identifying relatedness in regional systems.
Acknowledgements
Earlier drafts of the paper have been presented at workshops in Stavanger and Oslo. The authors are grateful to all participants at these events for their helpful comments and suggestions.
Disclosure statement
No potential conflict of interest was reported by the authors.
ORCID
Rune Dahl Fitjar http://orcid.org/0000-0001-5333-2701
Notes
1. Norway, with its population of 5.2 million inhabitants, is among the most sparsely populated countries in Europe. Consequently, economic regions are highly dispersed in both geographic size and population size, and the NUTS 3 regional classification is relatively large. Smaller Norwegian NUTS 3 regions are often the same size as NUTS 2 regions in other European countries, while regions located in the north are similar in geographical extent to the countries of Denmark, the Netherlands or Belgium. Hence, NUTS 4 regions is the best equivalent of functional economic regions. As we question the validity of these knowledge spillover dynamics in regions which cover very large areas, we rely on NUTS 4 regions in our analyses.
2. However, many economic regions are also represented by organizations such as regional councils or regional development agencies, which are normally joint ventures by several neighbouring municipalities with responsibilities especially for development policy.
3. The following economic regions are identified as part of the same labour market by Gundersen and Juvkam (Citation2013) and are therefore classified as one region: Oslo, Follo, Bærum/Asker, Lillestrøm and Ullensaker/Eidsvoll; Drammen and Sande/Svelvik; Tønsberg/Horten and Holmestrand; Skien/Porsgrunn and Kragerø; Kristiansand and Lillesand; Stavanger/Sandnes and Jæren; Haugesund and Søndre Sunnhordland; Trondheim and Stjørdalshalsen; and Namsos and Grong.
4. When calculating the regional skill relatedness index, industries are measured at the four-digit level in order to retain a reasonable number of employees in each industry. While the related variety index is measured at the five-digit level for consistency with Frenken et al.’s (Citation2007) approach, we have also tried calculating this at the four-digit level with no meaningful differences in the results.
5. Of course, labour mobility could be higher (or lower) between an industry pair in a particular region, leading these industries to be related in this region even if they are not related at the national level (or vice versa). However, the intention here is to identify industries which are related in a more general sense of building on the same skills, and we therefore calculate a national measure where industries are only considered related if the mobility levels between them are high across the whole country. It is thus not possible to determine from the relatedness maps whether there are actually high levels of mobility between any two industries in an individual region, only that these industries are related in the sense of building on the same skills as evidenced by higher than expected mobility at the national level.
6. The analysis focuses on employment in the private sector only. Following Frenken et al. (Citation2007), we further exclude the primary sector industries agriculture and fisheries from the analysis. We also exclude retail, hotels and restaurants, as well as temp agencies. The three former industries are excluded, as they include lots of temporary workers, while the latter is excluded, as it mainly acts as a channel to place workers in other industries; consequently, the knowledge and skills that are being transferred are not specific to the temp agency industry, but related to the industries in which temp agencies place workers.
7. The sum of all industries’ employment shares is by definition 1 in all regions. However, the root transformation implies that the sum of the weights will be higher for regions with more diversified industrial structures than for more specialized regions.