Abstract:

Based on the fifth and sixth national census data and the seventh national census bulletin of administrative regions at the prefecture level and above, the Gini coefficient, Theil index, spatial autocorrelation analysis and other methods are used to analyze the temporal and spatial evolution characteristics of talent agglomeration in Chinese cities from 2000 to 2020, and the multiscale geographically weighted regression is used to study the spatial heterogeneity and spatial scale of the influencing factors of talent agglomeration. The results show that: ①There is a large gap in the distribution of talents in eastern, central, and western cities in China, forming a distribution pattern of "strong in the east and weak in the central and west" and regional "one super and many strong", and the gap in average talent density tends to narrow. ②The distribution of talents shows great imbalance, with a slight weakening trend of instability. ③On the whole, the density of talents presents a significant positive spatial autocorrelation, and the spatial agglomeration effect of talents is becoming more and more significant. Locally, the Yangtze River Delta high-high talent agglomeration area and the Pearl River Delta high-high talent agglomeration area are gradually formed; the low-level talent agglomeration area composed of low-low talent agglomeration cities will have a homogeneous impact on its neighboring cities; some provincial capital cities in the west have significant high-low agglomeration characteristics, forming a "center-periphery" talent agglomeration pattern. ④The proportion of tertiary industry in GDP, the proportion of science and education expenditure in total fiscal expenditure, per capita GDP, the number of general higher education institutions, per capita public library collections, and the number of public transportation vehicles per 10 000 people have a significant positive impact on talent agglomeration, and the intensity of the impact decreases in order. Among them, the economic environment factor represented by the tertiary industry's share of GDP and per capita GDP has a relatively small spatial scale and strong spatial heterogeneity. Other independent variables are close to the global scale, and the spatial heterogeneity is not obvious.

Key words: talent agglomeration, temporal and spatial evolution, spatial autocorrelation analysis, Multiscale Geographically Weighted Regression (MGWR), spatial heterogeneity

摘要：

基于第五次、第六次全国人口普查数据和各地级及以上行政区第七次全国人口普查公报，采用基尼系数、泰尔指数、空间自相关分析等方法分析2000—2020年中国城市人才聚集的时空演化特征，采用多尺度地理加权回归研究人才聚集影响因素的空间异质性及空间尺度。结果表明：①中国东中西部城市人才分布密度存在较大差距，形成“东强中西弱”和区域性“一超多强”的分布格局，且平均人才密度差距具有缩小趋势。②人才分布表现出极大的不均衡性，具有非稳定的轻微减弱态势。③全局上，人才密度呈现显著的正空间自相关性，人才的空间集聚效应日趋显著。局部上，逐渐形成长三角和珠三角高-高型人才聚集区；由低-低型集聚城市组成的低水平人才聚集区会对其相邻城市产生同质化影响；西部部分省会城市具有显著的高-低型集聚特征，形成了“中心-外围”的人才聚集格局。④三产占比、科教支出占比、人均GDP、普通高等院校数量、人均拥有公共图书馆图书藏量、每万人拥有公共交通车辆数对人才聚集均有显著的正向影响，且影响强度依次降低，其中，以三产占比和人均GDP代表的经济环境因素空间尺度较小，具有较强的空间异质性，其他变量则接近全局尺度，空间异质性特征不明显。

关键词: 人才聚集, 时空演化, 空间自相关分析, 多尺度地理加权回归, 空间异质性