以往有关迁移流动影响生育水平的研究仍存在一个盲点，即缺少流动人口孩子出生地的分析。在弥补这个盲点后，分别通过泊松回归与Heckman二阶段模型再次分析了人口流动对生育水平的影响。泊松回归结果显示，农业户籍流动人口曾生子女数是农村本地人口的0.855倍，乡城流动人口的曾生子女数为农村本地人口的0.840倍，两类流动人口的累计生育率均低于农村本地人口。Heckman二阶段模型回归结果显示，农业户籍流动人口、乡城流动人口与农村本地人口的曾生子女数相比分别减少了0.237个与0.268个。因此，人口流动与生育水平降低确实有着因果影响关系。

This paper points out that a blind spot exists in the previous studies on the effect of migration mobility on fertility level, namely, the lack of analysis on the birth place of the children of the floating migration. After making up for the blind spot, Poisson regression and Heckman two-stage model are reused to analyze the effect of population mobility on fertility. The results of Poisson regression show that the number of the children of floating population with agricultural household registration is 0.855 times that of the rural locals, the number of the children of rural-urban floating population is 0.840 times that of the rural locals, and the cumulative fertility rate of the two types of floating population is lower than that of the rural locals. The regression results of Heckman two-stage model show that the number of the children of the floating population with agricultural household registration and that of rural-urban floating population reduce respectively 0.237 l and 0.268 compared with the number of children of rural locals. Hence, population mobility does exert a causal influence on the fertility decline.