What Drives the Settlement of Highly Educated Talents in China? A Social Integration Perspective

ABSTRACT In the current knowledge economy, regional strategies increasingly emphasize the attraction and retention of talent, making the settlement intentions of highly educated migrants a central concern for research and policymaking. While existing studies have largely examined explicit individual and regional factors, the influence of implicit dimensions, such as social integration, on long‐term settlement intentions remains insufficiently and systematically explored. To fill this gap, this study examines how social integration affects the settlement intentions of highly educated migrants in China. Drawing on 9202 highly educated respondents from the 2017 China Migrants Dynamic Survey, this study constructs a multidimensional index of social integration covering four dimensions: economic integration, social adaptation, behavioral adaptation, and identity recognition. It applies a binary logistic regression model, supplemented by endogeneity, robustness, and heterogeneity tests. The findings are fourfold. First, individual integration does not necessarily rise with regional economic development; in some developed advanced regions, high housing prices and living costs hinder integration. Second, regional factors such as economic development, geographic location, city size, and capital status are more strongly associated with settlement intentions than demographic characteristics or social integration. Third, among the dimensions of integration, identity recognition shows the most stable and salient association with settlement intentions, exceeding that of economic and social adaptation. Fourth, the relationship between integration and settlement intentions differs substantially across regions; in less developed areas, stronger economic integration is associated with greater mobility and lower local settlement intentions. Overall, this study deepens theoretical understanding of the relationship between social integration and settlement intentions, highlighting its contextual variations across regions. It identifies identity recognition as a core dimension and conceptualizes behavioral adaptation as a process variable. These findings offer practical implications for formulating differentiated regional talent policies.