Spatiotemporal Coordination between Human Capital Supply and Technological Innovation Actors

Addressing the skill–innovation coupling under industrial upgrading, this study leverages multi-source spatiotemporal data (2000–2023) on Shenzhen’s vocational training institutions (VTIs) and high-tech enterprises (HTEs) to develop an integrated framework—panel time-series, two-perspective nearest-neighbor proximity, information-theoretic spatial-entropy decomposition, and Geodetector—for identifying the patterns, mechanisms, and differences of multiscale spatial coordination. The findings are as follows. (i) The temporal evolution of vocational training follows a four-stage nonlinear transition; spatial structure expands from core areas toward corridors and the periphery, highly consistent with city development strategies and indicative of spatial path dependence. (ii) The two-perspective proximity analysis reveals widespread coupling at walking scale, alongside marked sectoral differentiation: scientific research/information and business services exhibit high-frequency, short-chain, nearby coordination, whereas finance shows central-place concentration and supply–demand asymmetry with weaker firm-side proximity. (iii) Spatial-entropy decomposition uncovers clear scale effects: slight co-location at 100 m, a shift to misalignment at 200–500 m, and polycentric dispersion at 1–5 km, yielding a nested structure of micro-scale coordination and macro-scale dispersion. (iv) At the district level, two typical regimes emerge: enterprise agglomeration with education outreach in core areas, and dual high entropies with small differences in peripheral new towns. Geodetector further indicates polarized functional association: pairings related to life services, scientific research, and information technology have significantly higher q-values, while manufacturing-related pairings are weaker and more uncertain. Accordingly, we propose a policy toolbox aligned with a “globally dispersed—locally coordinated” logic that combines multiscale, industry-customized, and locationally stratified measures: embed high-frequency training and apprenticeships within cores/corridors; enhance accessibility in peripheral zones via satellite/mobile sites and TOD; and mitigate polycentric separation through cross-area course/hour recognition and in-house-training spillovers—thereby improving the fine-grained coupling of industrial and talent chains and strengthening system resilience.