This study is motivated by the observation that, although universities have received increasing policy attention as key actors in regional innovation, empirical evidence remains limited in the Korean context on whether and how universities strengthen ...
This study is motivated by the observation that, although universities have received increasing policy attention as key actors in regional innovation, empirical evidence remains limited in the Korean context on whether and how universities strengthen the technological foundations of regional industries and foster technological diversification and innovation. In this context, the study examines the spatial characteristics of knowledge creation by Korean universities and investigates its relationship with regional technological specialization and diversification. It further aims to clarify the functions, limitations, and potential roles of Korean universities as actors within regional innovation systems. The findings can be summarized as follows.
First, technological knowledge creation in Korea operates within a capital region–centered structure dominated by the private sector. More than 75% of total R&D expenditure is carried out by the private sector, and over 80% is spatially concentrated in the Seoul Capital Area and Daejeon. University research funding relies heavily on the central government, which provides more than 70% of its financial sources, while more than half of such funding continues to be allocated to universities in the capital region, reinforcing a stratified and hierarchical structure. As a result, Korea’s R&D system is characterized by institutional and spatial concentration, and the knowledge-production structure of universities forms a pyramid-like hierarchy extending from the central government to elite capital-region universities, to non-capital flagship national universities, and then to other non-capital universities.
Second, analysis of the Weighted Average Inventor Distance (WAID) shows that universities with active knowledge creation, regardless of location, share the common feature of having their largest collaborative links with inventors within their own regions. At the same time, capital-region universities exhibit short average distances to inventors even when collaborating with multiple regions, indicating dense networks within close geographic proximity. In contrast, non-capital universities tend to collaborate across broader spatial ranges and display longer average distances, suggesting partial dependence on long-distance collaborations with the capital region and other metropolitan areas. Meanwhile, the Weighted Average Relatedness (WAR), which reflects the cohesiveness of university technology portfolios, remains close to the national average for most universities, implying broadly similar levels of technological cohesiveness under a central government–led research funding system. A quadrant analysis combining WAID and WAR classifies universities into four types: local cohesiveness, local variety, non-local cohesiveness, and non-local variety. Although non-capital universities are mainly distributed in non-local types, Kyungpook National University and Chonbuk National University appear as local-cohesiveness types, indicating that network forms can vary even among non-capital universities depending on regional industrial structures and the maturity of innovation systems.
Third, the study empirically demonstrates a dynamic relationship in which university knowledge creation does not precede regional technological diversification but instead tends to intensify after diversification entry occurs. System GMM results indicate that university patenting activity and regional technological specialization mutually reinforce each other over the long run, and that both regional technological structures and university knowledge creation exhibit strong path dependence and evolve through gradual and cumulative processes. These results suggest that technological entry and university research activity are not one-off events but co-evolutionary processes shaped through interaction.
Furthermore, the Linear Probability Model shows that regional technological relatedness density positively affects entry into new technological specialization, implying that diversification is more likely when existing local knowledge bases are closely related to the new domain. By contrast, lagged university patenting activity is negatively associated with new entry, suggesting that university research does not immediately translate into diversification. OLS results similarly show that university patenting significantly decreases in the year following new entry, but significantly increases in the same year as entry, indicating that university knowledge creation and regional diversification occur almost simultaneously. Finally, for technologies that sustain specialization for two consecutive years after entry, diversification events are found to increase subsequent university-involved patents. This implies that universities play a critical role in the stages of knowledge internalization and diffusion following new technological entry. In other words, rather than directly leading diversification, universities respond to industrial restructuring and contribute to strengthening emerging technological paths.
In conclusion, by analyzing the R&D activities of Korean universities, this study finds that universities function less as pioneers that open entirely new technological domains and more as knowledge intermediaries that recombine and diffuse knowledge within established technological systems, thereby dynamically sustaining and deepening the path-dependent evolution of regional industries. This pattern reflects Korea’s private firm-centered R&D structure and the institutional position of universities, which tends to respond to policy and industrial demand shifts rather than proactively setting new technological directions..
By incorporating Korea’s distinctive context of hierarchical stratification and capital-region concentration, the study contributes to broadening debates on regional technological specialization and diversification. From a policy perspective, it provides empirical evidence on the role of universities within Korea’s institutional and spatial context, offering implications for enhancing the effectiveness of regional innovation system policies such as the RISE programme.