Abstract: Using the spatial OD model，this study analyzes the migration flows among 31 provinces in China by exploring geographical effects，driving mechanism，and gender differences based on the inter-provincial migration data of the 6th Population Census． Inter-provincial migration exhibit significant spatial dependence，in which the origin and destination separate spatial autocorrelation promotes migration， whereas the origin-destination joint spatial autocorrelation hinders migration． In general， the “push”from origin is weaker than the “pull”from destination in Chinese inter-provincial migration．
There are significant gender differences in the driving mechanism． Impact of driving factors on female migration is stronger than that for male migration． Unemployment risk and employment discrimination are more evident in female migration． Males are more strongly influenced by spatial dependence than
females in the choice of migration destination， which could be explained by the gender differences of endowment and social environment．

Keywords: Migration Flows,  Geographical Effects ,  Spatial Dependence ,  Gender Differences ,  Spatial OD Model

摘要： 文章基于空间OD模型和“六普” 人口省际迁移流数据， 研究中国人口省际迁移流的地缘效应、 驱动机制与男女差异。文章得出的主要结论: ( 1) 中国人口迁移流内含显著的空间依赖，其中迁出、 迁入地独立的空间自相关效应促进人口迁移， 而迁出地-迁入地交互的空间自相关效应阻碍人口迁移。( 2) 总体上，中国人口省际迁移流受迁出地推力作用弱于迁入地的拉力作用， 表明人口迁移更多是出于对迁入地的“美好预期” ，而非对迁出地的“过度抱怨”。( 3) 人口迁移流的驱动机制有显著的性别差异: 女性迁移流受驱动因子的影响强于男性，女性的失业风险、 就业歧视更为明显; 男性在人口迁移的空间选择上受空间依赖的影响强于女性，这与性别禀赋和社会环境有关。

关键词: 人口迁移流, 地缘效应 , 空间依赖 , 性别差异 , 空间 OD 模型