Bayesian Dynamic Analysis on Innovation Adjustment to SME Human Capital Capacity in EU Regions

Abstract This paper examines how SME human capital capacity shapes the dynamic evolution of regional innovation performance across European Union NUTS2 regions between 2018 and 2025. Using data from the Regional Innovation Scoreboard and Eurostat, we construct an SME human capital capacity index that combines tertiary education, lifelong learning, ICT specialists, and HR-intensive SME innovation indicators such as innovative SME employment and collaborative innovation. We then estimate a dynamic Bayesian panel model with a lagged dependent variable to quantify persistence in regional innovation and the interaction between business R&D and SME human capital capacity. The results show that regional innovation is highly persistent, and that SME human capital capacity strongly amplifies the short-run and long-run effects of business R&D on innovation performance. Regions with a stronger SME human-resource base exhibit faster adjustment to shocks and higher steady-state innovation levels, emphasising the central role of skills, learning, and HR-intensive innovation in regional development strategies. JEL codes: O31, O32, L11.