1995—2015年中国省际人口迁移网络的演化特征——基于异质性劳动力视角

自户籍制度放宽以来,中国经历了大规模的人口迁移。其中,异质性劳动力（高技能劳动力、普通劳动力）在迁移过程中表征出不同的空间格局和网络组织特征,对地区经济发展产生各异的影响。本研究基于人口普查和抽样调查微观抽样数据,使用复杂网络理论对1995—2015年中国异质性劳动力迁移的时空格局和网络演化展开分析,并分析了格局背后的可能成因。研究发现：① 省际高技能和普通劳动力迁移均呈现出持续高度不平衡的空间集聚特征,承载大量人口的迁移流主要由中国中西部地区指向东部沿海地区,但这种空间不平衡特征呈现出一定的减弱趋势。高技能劳动力的平均迁移距离比普通劳动力更长;② 两类劳动力迁移网络中均呈现明显的“小世界”特性,但普通劳动力迁移网络的迁移强度和关联程度均高于高技能劳动力迁移网络;③ 两类劳动力迁移网络均呈现出以北京、上海、广东为主要核心节点的网络结构。高技能劳动力网络结构呈现相对稳定的特征,而普通劳动力网络结构则呈现出变化的趋势;④ 地区经济差异、路径依赖效应、异质性劳动力在就业市场上的差异性以及劳动力市场对异质性劳动力需求的差异是导致两类劳动力迁移格局差异性的重要成因。

Evolution patterns of China's interprovincial migration networks between 1995 and 2015:Based on labour heterogeneity

With the loosing of the household registration (hukou) system, China has witnessed massive volumes of interregional migration. Among all kinds of migration, skilled migration and less-skilled migration present differential spatial patterns and network characteristics over time and may exert distinguished impacts on regional development and innovation. However, due to limited data availability, prior studies only adopted cross-sectional data analysis and examined the patterns of skilled and less-skilled migrations before 2005, whereas few have been done on related topics after 2005. Drawing on previous studies, the present paper uses micro-level datasets from population censuses and sample surveys to assess the spatial evolution patterns of China's skilled and less-skilled internal migrations between 1995 and 2015. Besides, several complex network analysis methods have been employed, together with the network visualization technique. The results of the paper are listed as follows. First, there were spatial clustering and uneven patterns in both skilled and less-skilled internal migrations over the 20 years of 1995-2015, where migration flows containing a large percentage of skilled and less-skilled migrants originated from inland western and central regions to the eastern coastal areas of China. However, the degree of spatial unevenness in the migration patterns has been slightly weakened. In addition, the average migration distance of skilled migration was farther than that of less-skilled migration in the same period. Second, both of the two migration networks have presented a “small-world” property over the 20 years. In comparison, the intensity and connection degree of less-skilled migration networks have prevailed over that of the skilled migration networks. Third, results from the community detection have revealed that both skilled and less-skilled migrations exhibited a network structure where Beijing, Shanghai, and Guangdong served as the core nodes. Skilled migration presented a more stable network structure during the 20 years, while less-skilled migrations showed a persistently changing pattern. Fourth, the paper suggested that regional economic disparity, path dependence effect, the heterogeneity of different labour forces in the job market, and the differences in labour market demand for different labour forces were underpinning factors shaping the spatial evolution pattern and network structure of the two types of migration from 1995 to 2015.