In the context of China's "dual-carbon" goals (carbon peaking and carbon neutrality) and the energy transition, the new energy vehicle industry has exhibited pronounced spatial agglomeration under the joint influence of policy support and technologica...
In the context of China's "dual-carbon" goals (carbon peaking and carbon neutrality) and the energy transition, the new energy vehicle industry has exhibited pronounced spatial agglomeration under the joint influence of policy support and technological progress. Such clustering can reshape regional economic performance through factor mobility and network-based collaboration, with spillovers that may manifest as either positive diffusion or negative crowding-out. However, existing studies lack an integrated framework linking cluster measurement, spatial-effect identification, and mechanism testing, which limits assessment across alternative spatial specifications and obscures the channels through which cluster effects operate. This study investigates how new energy vehicle industrial clusters affect economic growth using provincial panel data for China over 2011-2022. Economic performance is measured by CPI-adjusted GDP per capita in logarithms, while cluster intensity is proxied by the location quotient (LQ). The analysis controls for capital, labor, science and technology input, infrastructure, and human capital. The baseline specification is a Spatial Durbin Model (SDM) with two-way fixed effects, and spatial dependence is first confirmed using Moran’s I. To assess the sensitivity of spatial linkages, the study estimates the model under an economic-distance spatial weights matrix and an economic – geographic composite matrix, and interprets results through the decomposition into direct, indirect (spillover), and total effects. Robustness checks are conducted by replacing the spatial weights matrix and substituting alternative dependent variables. To address potential endogeneity, the study applies an instrumental-variable two-stage least squares approach (IV– 2SLS) and evaluates exogeneity using the Durbin–Wu–Hausman test. Finally, mechanism tests examine the transmission channels of cluster effects by using the number of NEV-related patents (RD) as a proxy for technological innovation diffusion and the number of related firms (Indu) as a proxy for industrial-chain collaboration, incorporating interaction terms and spatially lagged variables to identify the pathways of spillovers. Empirical results indicate that NEV industrial clusters promote regional economic performance, but the estimated spillover pattern depends on the spatial weights matrix. Under the economic-distance matrix, the decomposed direct, indirect (spillover), and total effects are all positive but not statistically significant, implying that the direction of influence is consistent while the evidence is not strong enough to draw a definitive conclusion on spillovers among economically similar regions. By contrast, when geographic proximity is incorporated through the economic – geographic composite matrix, the indirect effect becomes significantly positive and larger than the direct effect, and the total effect is also significantly positive. This suggests that diffusion effects are more likely to be realized when regions are geographically adjacent and connected through practical channels such as logistics, supply-chain linkages, and interregional interaction. The mechanism analysis provides stronger support for the technological-innovation channel. Patent output significantly promotes regional growth, and once innovation is controlled for, the direct effect of clustering weakens while the cluster–patent interaction term remains significantly positive. In contrast, evidence for the industrial- chain collaboration channel is comparatively weaker and less stable. The number of related firms does not translate mechanically into growth gains, and the cluster– industry-chain interaction term is significantly negative in the within-region (direct- effect) estimates, suggesting that expanding related firms may dilute the marginal growth contribution of clustering when it is associated with homogeneous expansion or resource crowding. Spillover reinforcement through supply-chain networks therefore appears to be conditional on the quality of interregional specialization and coordination rather than on a simple increase in firm counts. This study proposes an integrated framework linking cluster measurement, spatial effects, and mechanisms. Using a Spatial Durbin panel model, it estimates the direct and spillover effects of NEV clusters on regional growth and incorporates strategic emerging industries' spatial externalities into agglomeration theory. The findings show a dual effect: clusters can promote coordinated growth via innovation networks but may reduce gains among similarly developed regions under intensified competition. The study advances the cluster-spillover literature and provides evidence for cluster- oriented policies in policy-driven high-technology industries. The policy implications are threefold. First, NEV cluster policy should prioritize interregional linkages and diffusion rather than local scale expansion alone; policy evaluation should focus on whether clustering promotes coordinated regional growth and yields sustainable spillovers. Second, positive spillovers require stronger cross-regional infrastructure—especially transport and logistics — fewer institutional and market barriers, and lower collaboration costs through unified standards, coordinated regulation, and benefit- sharing mechanisms; meanwhile, governments should avoid duplicative investment and homogeneous industrial planning by clarifying the division of labor and leveraging regional complementarities. Third, while supporting leading regions in strengthening core innovation capacity, policymakers should facilitate the diffusion of experience and technology to latecomer regions via technology-transfer platforms, interregional R&D collaboration, and outcome sharing. With greater factor mobility and cooperation, NEV clusters can drive coordinated regional development rather than intensify competition. Keywords: new energy vehicle (NEV) industrial clusters; regional economic growth; spatial spillover effects; Spatial Durbin Model (SDM); technology diffusion and industrial chain collaboration