According to previous studies, not only does the conditional gravity model based on ordinary least squares often bring about poor fitting of migration flows in reality, but also there exists overdispersion in the extended Poisson gravity model. Simultaneously, network autocorrelation usually exists in population migration data （e.g., the spatial interaction among migration flows). The problems mentioned above result in biased estimation. In order to capture network autocorrelation and deal with the issue of overdispersion, we build an eigenvector spatial filtering negative binomial gravity model (ESF NBGM) based on the data of 1% national population sample survey in 2015, to analyze the driving mechanism of interprovincial population migration flows in China. The results are as follows: (1) Positive spatial spillover effect exists in interprovincial population migration flows, and ESF can capture network autocorrelation in data, so as to reduce the estimated deviation of the model. Furthermore, eigenvectors ranking top 1.4% can properly interpret the spatial pattern of high network autocorrelation in data. (2) There exists overdispersion in China's interprovincial migration flows. Considering this problem, a negative binomial regression model is more suitable for the estimation of driving mechanism for population migration, together with statistical enhancement. (3) Network autocorrelation leads to overestimation of distance variables and underestimation of non-distance variables. The results of the improved model reveal that: chief factors the affect driving mechanism are regional population characters, social network, economic development and education level. Meanwhile, living environment and road network gradually become one of the most crucial pulling factors that influence migration flows. (4) Compared to previous studies, social network (i.e. migration stock) plays a more significant role in population migration flows, while the impact of spatial distance keeps weakening.

人口迁移数据中往往存在较强的网络自相关性,以往基于最小二乘估计的重力模型与迁移数据的拟合度较低,而改进后的泊松重力模型仍存在过度离散的缺陷,以上问题均导致既有人口迁移模型中的估计偏差。本文构建了特征向量空间滤波（ESF）负二项重力模型,基于2015年全国1%人口抽样调查数据,研究2010-2015年中国省际人口迁移的驱动因素。结果表明：① 省际人口迁移流间存在显著的空间溢出效应,ESF能有效地提取数据中的网络自相关性以降低模型的估计偏差,排序在前1.4%的特征向量即可提取较强的网络自相关信息。② 省际人口迁移流之间存在明显的过度离散现象,考虑到数据离散的负二项重力模型更适用于人口迁移驱动因素的估计。③ 网络自相关性会导致模型对距离相关变量估计的上偏与大部分非距离变量估计的下偏,修正后的模型揭示出以下驱动因素：区域人口特征、社会网络、经济发展、教育水平等因素是引发省际人口迁移的重要原因,而居住环境与公路网络等因素也逐渐成为影响人口迁移重要的“拉力”因素。④ 与既有研究相比,社会网络因素(迁移存量、流动链指数)对人口迁移的影响日益增强,而空间距离对人口迁移的影响进一步呈现弱化趋势。

Talents are the core driving force for regional innovation and development, and they are also the key to urban competition in the future. Based on the data of the sixth population census and the national 1% population sampling survey in 2015, this paper uses the concentration index (CI), Gini index (GI) and several related spatial analysis methods to examine the spatial patterns of highly educated talents across China and the drivers behind these patterns from 2010 to 2015. The results are as follows: (1) Talents show highly concentrated and unbalanced spatial distribution patterns at the city level over the five years, but the trend of concentration and imbalance has gradually eased. Results from standard deviation ellipse (SDE) indicate that the distribution direction of highly educated people is "NE-SW". Furthermore, there is a significant spatial spillover effect in the distribution of talents, with three major urban agglomerations of China as the highlands. (2) Economic opportunities are the dominant drivers for the distribution of highly educated talents in China. Among them, wages are the core driver, and gross GDP and industrial structure of each city also exert a significant impact. (3) After controlling the impacts of economic opportunities, local quality variables represented by basic public services (e.g., education and medical care), transportation accessibility and urban consumption facilities play an important role in the distribution of highly educated talents. (4) There are significant differences between the driving factors for talents in urban agglomerations and non-urban agglomerations: economic opportunities are the main driving force for the distribution of talents in both urban agglomerations and non-urban agglomerations, while local qualities including education, consumption, transportation and natural environment have a more significant impact on the distribution of talents in urban agglomerations. This study provides references for the formulation of urban and regional talent policies and regional development policies.

人才是推动地区创新发展的核心动力,也是未来城市竞争的关键。基于第六次全国人口普查和2015年全国1%人口抽样调查数据,运用集聚度、基尼系数等空间统计分析以及负二项回归模型方法,对2010—2015年中国城市高学历人才的空间演化特征及驱动因素展开研究。结果发现：① 人才呈现出高度集中且不平衡的空间分布格局,但不平衡的趋势有所缓解,分布方向为“东北—西南”走向。人才分布的空间溢出效应显著,城市群是人才集聚的高地。② 经济机会是影响中国高学历人才集聚的主导力量,其中工资是核心驱动因素,地区的发展水平、产业结构同样产生显著影响。③ 控制经济机会的影响后,以教育和医疗等基础公共服务、交通可达性和城市消费设施为核心的地方品质变量在高学历人才的空间集聚过程中发挥重要作用。④ 城市群与非城市群人才驱动因素存在显著差异：经济机会是城市群和非城市群人才集聚的主要驱动力;教育、消费、交通以及自然环境等地方品质变量对城市群人才集聚的影响更为显著。本研究为城市和区域人才政策和区域发展政策的制定提供借鉴与参考。

"In this paper, [the authors] suggest an alternative method for fitting the gravity model. In this method, the interaction variable is treated as the outcome of a discrete probability process, whose mean is a function of the size and distance variables. This treatment seems appropriate when the dependent variable represents a count of the number of items (people, vehicles, shipments) moving from one place to another. It would seem to have special advantages where there are some pairs of places between which few items move. The argument will be illustrated with reference to data on the numbers of migrants moving in 1970-1971 between pairs of the 126 labor market areas defined for Great Britain...."excerpt

By looking into the enrollment and employment data of graduates from universities directly administered by China Ministry of Education and with the method of index evaluation and hot-cold spot analysis, this article analyzed the graduates’ regional flowing patterns of two flowing stages of enrolled in the university and employed after graduation, and the resulting spatial distribution at the provincial level. The study found that the flow of talent from the university to study in different stages, can more clearly reveal the characteristics of the geographical space for talents. The local spatial viscosity in different regions dominates the flow of talent, and geopolitical and income factors in subsequent plays a role of regional adjustment. In the two stages, the flow of talent has significant spatial viscous characteristics. The geographical pattern of the flow includes “local-leapfrog” mode, “local-semi adherent” mode and “local-adherent” mode. Under the influence of different factors, the enrollment stage is dominated by “local-(semi) adherent” mode due to the adherence to the geo-social relations, while the employment stage is dominated by “local-leapfrog” type, which is adhered to the multiple possibilities of regional employment opportunities and benefits. From the perspective of the provincial pattern formed by talent flow, however, the spatial distribution of talents at the level of provincial level is more flat, while the phase of employment flow is more polarized in the longitudinal distribution. The “arch” pattern along the southeastern coast and the Yangtze River is characterized in both two stages, and the Yangtze River Delta region belongs to the hot spot of talent. However, because of the lack of provincial integration and linkage, the centralization of talent is not significant enough in the central and western regions, which highlight the important effect of the dominant area’s viscosity in the formation of the high ground of talent. It is suggested that different cities should bring into full play the role of local glutinosity to enhance the work of introducing university intelligence, from the two stages of talent generation and with the help of the strength of the urban agglomeration.

依据教育部直属高校2015届本科毕业生生源与就业数据,采用指标评价与冷热点分析方法,分析升读大学与本科就业两个流动阶段的人才流动地域模式及省域空间分布格局。研究表明：① 人才流动具有明显的本地空间粘滞性特征,地域模式包括“本地-跃迁”型、“本地-半依附”型和“本地-依附”型。② 省际层面形成沿东南沿海与长江沿岸分布的“弓形”格局,显示了优势区域的整体粘滞性对人才高地形成的意义。研究指出,把握关键节点、依托城市群来发挥粘滞作用有助于城市推进引智工作。

As one of the important contents of urban geography, regional urban spatial organization is turning from the traditional city system research to the city networks research. There are two methods to construct city networks. One is using the relational data including the goods flow, capital flow, information flow, population flow among the city to construct city networks; the other one is building models to simulate the relationship between the city, such as interlocking network model and the gravity model. Inspired by the view of the "a spatial dynamic that changes whenever actors move in space" proposed by Menzel, this paper attempts to construct the city networks by using returnees' movement between cities. First, this paper discusses the theoretical relationship between talent mobility and city network. Relational economic geography uses the idea of "based on construction of individual relationship " into network analysis, evolutionary economic geography puts forward the thought of "dynamic proximity", and the research of early uses mobility of talents of high technology to measure the relationship between cities. These had laid a theoretical foundation for the construction of the city network based on the flow of talents movement. Then, we establish the conceptual framework of "triangle model of Chinese returnees". The returnees have the fellow villager's relationship resources in their native places, the classmates' relationship resources in overseas cities and the coworkers' relationship resources in the work city, and these are used in the analysis framework. By integrating the migration data of the returnees, we establish the cities' connections based on the flow of talent, and design a method to calculate the city connectivity. Finally, we select 6 high-end returnees of an innovative team in China to carry out a case analysis. Through the mining of the returnees' city resumes, we successfully construct the city network composed of six talents. As a result, it is suggested that: (1) city innovation network should be examined at national scale under global perspective, and the returnees are good carriers in such researches; (2) the "triangular model" reflects Chinese characteristics, which can explain the formation mechanism of city innovation network better; (3) the returnees team in Beta Pharma Co., Ltd is a typical case that it successfully facilitates inter-city innovation cooperation; and the "triangular model" framework provides a brand-new perspective and analytical method to further research on city innovation network. It should be noted that this paper focuses on theory construction and method exploration, and it needs to use a great quantity returnees' data to carry out systematic research on urban network in the next step, in order to make a greater contribution to urban network theory and the strategy of talents creative practice.

以高端归国人才为对象,通过构建“归国人才三角”概念框架,设计基于高端移动性人才的城市关系构建方法,对人才迁移驱动的城市网络构建进行了初步探讨。研究认为：① 借助“行动者空间移动带来的空间动力”这一视角,运用高端归国人才在城市间的移动模拟城市间关系,可为构建城市网络提供一条新路径。② 基于人才流动的城市网络构建,有助于破解城市关系矩阵数据挖掘与案例城市选取的难题,是知识经济时代城市网络研究的新方向。③ “归国人才三角”概念框架将人才流动与城市网络两方面研究融为一体,既可以通过人才流动探讨城市创新空间组织特征,也可以通过城市网络方法分析人才流动网络,对城市网络研究和人才战略实施都具有一定参考价值。

基于2000和2010年全国人口普查分县数据和地级行政单元数据,采用不均衡指数和空间自相关分析等方法,刻画中国高技能劳动力和一般劳动力的空间分布格局及其变化,并采用空间滞后模型,识别高技能劳动力与一般劳动力空间分布的决定因素及其变化。结果表明：① 两类劳动力总体的空间特征均为东南密集,西北稀疏;② 2000—2010年,高技能劳动力在空间分布上呈集中化的趋势,而一般劳动力呈分散化的趋势;③ 十年间高技能劳动力集聚区虹吸作用加强,其分布越发集中于沿海特大城市群,而一般劳动力集聚区回波作用加强,其分布越发均衡;④ 空间回归分析结果表明,职工平均工资、是否直辖市或省会、每万人在校高校学生、是否沿海城市、常住人口数和空间溢出效应在十年中一直都是决定高技能劳动力和一般劳动力空间分布的主要因素,而失业率、中学生师比、二氧化硫排放量和绿地率在2010年也成为影响两类劳动力空间分布的主要因素。对比两类劳动力的模型结果可得,高技能劳动力的空间分布受行政因素、高校因素和地区生活舒适度的影响更大,而一般劳动力的空间分布受劳动力市场因素的影响更大。

Academicians, as national high-end research leaders, play an important role in innovative development, talent training, scientific research guidance, national major strategies and core technology breakthroughs. Based on the survey of 755 academicians of the Chinese Academy of Sciences (CAS), the spatial analysis methods such as nuclear density analysis and mobility index are used in this paper to systematically analyze the characteristics of spatial distribution and three flow patterns of study, employment and changing workplaces of China's top academic talents. The results indicate that: (1) The birthplaces or native places of the academicians are mainly located in the eastern coastal regions with good economic conditions and profound cultural heritages; The academic qualifications of the academicians are mainly obtained in regions such as Beijing, Shanghai and Jiangsu Province, where top universities (also known as 985 and 211 universities) are located. The first workplaces of the academicians present different manifestations in different periods. The final workplaces are mainly distributed in megacities such as Beijing and Shanghai. (2) The flow of education (birthplace-university) moves from some provinces in the east to the regional center cities, such as Beijing and Shanghai. After graduation, the employment flow shows the characteristics of a large number of destinations which vary with different periods of different academicians, and the degree of spatial aggregation is small. (3) Some 46.23% of the academicians have a tendency of changing their workplaces. The flow of academicians who changed their workplaces one time presented the agglomeration characteristics of “peacocks fly to southeast”, which comes from a long narrative poem with the same name in ancient China, this article uses it to describe the phenomenon that many China's top academic talents moves to the southeast coastal zones, with obvious spatial orientation; the academicians who changed their workplaces two or more times have no apparent spatial aggregation feature except for the last flow by the cause of changing workplace. (4) The analysis of flow index shows that Beijing and Shanghai, as rapidly developed economic zones, are both strong inflow and strong outflow centers with large two-way flow; while Tibet, Hainan, Macao, Inner Mongolia, and Guangxi provincial-level regions have a zero mobility index, and they are subordinate provinces.

院士作为国家高端的科研领军人才,在创新发展、人才培养,科研引领,国家重大战略和核心技术突破方面挥着重要作用。以中国科学院755名院士为例,运用核密度分析与流动指数等空间分析方法,系统分析了中国顶尖学术型人才的空间分布特征及求学、就业与工作地变更三种空间流动规律。结果表明：① 院士出生地或籍贯集聚分布在经济条件良好、文化底蕴深厚的东部沿海地区;学历取得地主要集中在北京、上海与江苏等985、211高校聚集省区;初次工作地点在不同时代表现各异;最终工作地呈现向北京、上海等特大城市聚集的特征。② 求学（出生地—大学）流动由东部多个省份向北京、上海等国家中心城市近距离流动。毕业首次就业流动空间聚集程度小,且流向随院士所处时代的不同而变化。③ 46.23%的院士有工作地变更流动趋势。变更一次的院士其流动呈现“孔雀东南飞”的特点,具有明显的向东指向性和空间聚集性;变更两次及两次以上的院士,除最后一次流动外无明显空间指向性。④ 流动指数分析表明,北京与上海作为经济发达区既是强流入中心也是强流出中心,双向流动量大;而西藏、海南、澳门、内蒙古及广西等省份流动指数为0,属于附属型省份。