基于空间计量交互模型的人才流动影响因素研究——以中国“双一流”高校毕业生为例

China has entered the stage of transformation from a "demographic dividend" to a "talent dividend", and talent has increasingly become the key driver of national and regional development in the era of a knowledge economy. As the reserve of talents, university graduates are the main target of the "war for talent" among Chinses cities. Based on the 2019 Graduate Employment Quality Reports of China's "double first- class" universities, adopting the Gini coefficient, spatial autocorrelation, and spatial econometric interaction model, this paper demonstrates the migration pattern of "double first- class" university graduates upon their graduation and investigates its underlying influencing factors. The results reveal that the destination areas of university graduates from "first- class universities" and "first- class disciplines" are highly concentrated in eastern China, with the former showing a higher concentration level. While economic factors still play a vital role in determining the migration of university graduates, the influence of quality of place is also significant, especially for graduates from "first- class universities". In addition, the policy factors also significantly influence the migration pattern of graduates from both types of universities. There are significant network autocorrelation effects among graduates' employment migration flows. The network autocorrelation effects based on places of study and places of employment are significantly positive. These network autocorrection effects reinforce the uneven distribution pattern of university graduates' migration. This study highlights the importance of employing a regional coordination perspective rather than a single- region perspective in terms of the formation and further optimization of regional talent policies.