How specialization and diversity of knowledge base contribute to regional knowledge complexity

Why can complex knowledge be produced in some places but not in others? Regional knowledge complexity possesses a powerful explanatory and predictive capacity for regional innovation and economic development, yet the mechanisms underlying the formation of regional knowledge complexity remain largely opaque. This study aims to reveal the how specialization, related variety and unrelated variety of regional knowledge base contributing to the formation of regional knowledge complexity, which enriches the theories in recombination innovation as well as regional complex system. We employed multiple regression models to analyze how specialization, related variety, and unrelated variety within the regional knowledge base influence the regional Knowledge Complexity Index (KCI). The research findings indicate that specialization exerts a positive impact on the regional KCI, and the related variety within a regional knowledge base positively impacts KCI but only under the condition of cognitive proximity between regions. Both specialization and Regional Variety kjm(RV) exert significant externalities: a one-unit rise in specialization locally enhances a city's KCI by 0.0002 and affects neighboring cities positively by 0.0414, while a similar increment in RV elevates the host city's KCI by 0.0279 and induces an 8.5022-unit jump in nearby cities' KCI. Conversely, the unrelated variety in a regional knowledge base has a negative effect on KCI. A one-unit increase in the unrelated diversity of a region's knowledge composition correlates with a decline of 0.0147 units in the city's KCI and that of 2.0507 units in neighboring city's KCI. The findings suggest that less-developed regions should allocate their limited resources judiciously by prioritizing the development of specialized advantages, and concurrently, work towards narrowing the cognitive gap with their more advanced counterparts to facilitate the flow of knowledge.