Skills tend to be mysterious entities in economic geography and regional science – notoriously difficult to define, identify and measure. Yet, almost everyone agrees on their decisive value for regional economic development and welfare. In our new, freely available study, we propose that large-scale data on job advertisements represent a substantial and largely untapped opportunity to revisit old findings and assumptions, to overcome bottlenecks in existing empirical approaches, to pose new questions and to inform policy in relation to regional research on skills.
We tested this conjecture in the context of Sweden, where combining job advertisement data with register data allowed us to study the skill-similarity of jobs (industry-occupation combinations), the coherence of local skill portfolios, and regional development outcomes at a granular spatial scale.
First, the microscope…
We introduced a job-postings-based approach to measuring the similarity of skills required by employers across jobs. Previous approaches relied on labour mobility between economic activities to reveal this skill-relatedness, assuming that workers tend to switch between jobs where they can reuse much of their accumulated experience and skills. Our approach is not directly dependent on labour mobility, thereby providing a scalable and potentially region-specific alternative to traditional labour flow-based approaches, allowing for a better understanding of local skill demand and its evolution.
We showed that workers tend to be more likely to switch between jobs with similar skill-demand, thereby offering direct support for this assumption behind revealed skill-relatedness. An openly available job-job skill-similarity table accompanies the paper. The skill-relatedness of jobs shows strong clustering both in terms of occupations (Figure 1A) and industries (Figure 1B), attesting to the specificity of human capital to economic activities.
Figure 1. The job-based skill-relatedness space represented as a network. (A) Nodes are coloured according to occupations; (B) nodes are coloured according to industries.
… then some exploration
While most Swedish municipalities exhibit above-average coherence in terms of skills, less coherent regions, with more diverse skill structures, tend to be linked to higher education levels and economic performance. This highlights the potential benefits of unrelated variety and diversity in regional labour markets, especially in cities like Stockholm, Malmö, or Gothenburg.
We also find that regions with higher rates of skill-related job switches tend to have better economic outcomes. This highlights the importance of skill transferability and the role of dynamic labour markets in regional development.
Henning, M., Eriksson, R., Garefelt, P., Martin, H., & Elekes Z. (2025): Job relatedness, local skill coherence and economic performance: a job postings approach. Regional Studies, Regional Science, 12(1): 95–122.
