International Entrepreneurship and Management Journal Управление технологиями и инновациями

Technology transfer and innovation dynamics through clustering analysis of European Union countries

Zherlitsyn, Dmytro, Kolarov, Kostadin
Springer Nature · 2026-10-21 · DOI 10.1007/s11365-026-01259-w

Аннотация

Despite uniform EU innovation policies, countries persistently differ in R&D investment levels, firm-level digital adoption, and the ability to translate aggregate research spending into enterprise-level outcomes. This study explores the diversity in innovation performance and technology transfer capacity among EU member states, with particular attention to small and medium-sized enterprises. The study applies a three-level clustering framework to Eurostat panel data. Clustering 1 characterises macro-level innovation capacity using average data from 2015 to 2024 of GDP per capita, GERD per inhabitant, and R&D intensity as a share of GDP. Clustering 2 compares country positions at two fixed endpoints, 2015 and 2023, to identify structural transitions over the decade. Clustering 3 uses 2023 Eurostat size-class aggregates for enterprises with 10–249 persons employed, characterising enterprise-level technology transfer capacity through two indicators: business R&D expenditure as a share of SME value added and SME value added per employee. Hierarchical agglomerative clustering with Ward's linkage is applied at each level, and all solutions are validated using the Silhouette Score and the Adjusted Rand Index. The analysis identifies five macro-level innovation tiers and six SME-level profiles. Cross-level comparison reveals that macro-level R&D leadership does not automatically produce firm-level technology transfer outcomes. Germany, for example, ranks fourth in the EU on business R&D expenditure, yet clusters with Cyprus and Malta at the SME level. For Bulgaria and similar economies, the findings suggest that institutional configuration mediates the effectiveness of technology transfer more strongly than aggregate spending levels alone. The results support a differentiated rather than uniform approach to EU cohesion and innovation policy design.

Читать у издателя →