Talent agglomeration and carbon emissions in China: promotive effect or inhibitory effect? analysis on heterogeneous bilateral stochastic frontier model

As the dual imperatives of carbon reduction and economic expansion gain prominence, the pursuit of high-quality development has taken on greater urgency, with talent agglomeration emerging as a critical driving force. Yet its specific impact on carbon emissions remains unclear. To address this question, this study employs a two-tier stochastic frontier model using panel data from 30 Chinese provinces spanning 2013–2023, enabling the identification and decomposition of the dual effects of talent agglomeration on carbon emissions. The results show that talent agglomeration exerts asymmetric bilateral impacts, with the inhibitory effect dominating the promotive effect and producing an overall decline in emissions. Specifically, the promotive channel increased emissions by 16.2%, while the inhibitory channel reduced them by 43.8%, resulting in a net decrease of 27.6% relative to the frontier level. The net inhibitory effect became more pronounced after 2020 following a brief uptick in 2014, and spatial heterogeneity was evident, with the largest reductions occurring in the western region, followed by the central and eastern regions. Further heterogeneity analysis indicates that larger net reductions occur at higher levels of talent agglomeration (≥40%), higher economic development (70%–100%), and under stricter environmental regulation, whereas in the early stages of urbanization the promotive effect dominates but gradually shifts toward inhibition as urbanization deepens. These findings demonstrate that although talent agglomeration has dual channels of influence, its overall impact is to reduce emissions, thereby providing more precise guidance for carbon reduction policies that reconcile economic growth with environmental sustainability.