Technological Proximity and Dyadic Urban Coordination: Nonlinear Dynamics Across Chinese Cities

How technological proximity shapes inter-city coordination remains insufficiently understood, particularly when proximity may generate both learning gains and congestion costs. Using inter-city patent transfer records for 253 Chinese prefecture-level cities from 2014 to 2023 and more than 65,000 directed city-pair observations, this study examines the relationship between technological proximity and dyadic urban coordination. We first construct a multidimensional city-level development score using the entropy-weight method and then derive a city-pair Urban Coordination Index through a Coupling Coordination Degree (CCD) model. Fixed-effects panel estimations with a quadratic specification reveal a robust inverted U-shaped relationship: dyadic urban coordination initially improves as technological proximity lowers learning barriers, strengthens absorptive alignment, and facilitates knowledge recombination, but declines once excessive similarity generates redundancy, congestion, and lock-in pressures. Further analyses show that industrial similarity and patent flow intensity both steepen this inverted U-shaped relationship, indicating that stronger structural overlap and more intensive knowledge exchange bring forward the point at which additional proximity becomes counterproductive. Heterogeneity analyses further suggest that the downturn is sharper in more developed regions and in center–center dyads, whereas noncenter dyads exhibit weaker gains, consistent with thinner innovation infrastructures and more limited system capacity. These findings clarify the nonlinear dynamics through which technological proximity shapes inter-city coordination and highlight the structural and regional conditions under which proximity supports or undermines coordinated urban development.