Abstract: The under-reporting of infant death undermines the accuracy of the infant mortality rate and life expectancy. Based on China’s 2010 national population census data, this paper analyzes the impact of infant death under-reporting on life expectancy from both theoretical and empirical perspectives. Cluster analysis and spatial autocorrelation analysis have been conducted to explore the regional disparities of both the under-reporting rate of infant death and its influence on life expectancy. The results show that the lower the overall mortality rate is, the more the change of life expectancy caused by the change of infant mortality. The extent of the impact on life expectancy of infant mortality depends mostly on its under-reporting level, the higher the level, the larger the impact. There are significant regional disparities of the impact of infant death under-reporting on life expectancy. The life expectancy is more sensitive to the fluctuation of infant death under-reporting in western and southwestern China. Moreover, there is no global bivariate spatial correlation between mortality level and the likelihood index of infant death under-reporting, but there are some significant local agglomerations and the spatial patterns are similar by gender.

Keywords: Infant Death Under-reporting, Life Expectancy, Likelihood Index of Infant Death Under-reporting, Cluster Analysis, Spatial Autocorrelation

摘要： 婴儿死亡漏报是影响婴儿死亡率准确性及平均预期寿命可信度的重要因素。文章基于2010年中国人口普查数据，从理论和实证两个方面分析了婴儿死亡漏报对平均预期寿命的影响，并利用聚类分析和空间自相关分析探究了婴儿死亡漏报率的可能水平及其对平均预期寿命的影响的地区差异。研究发现死亡水平越低，婴儿死亡率变动所引起的平均预期寿命的变动就越大；婴儿死亡漏报率越高，婴儿死亡漏报率的变动所引起的平均预期寿命的变动就越大；婴儿死亡漏报率变动对于平均预期寿命的影响存在显著的区域差异，西部和西南地区平均预期寿命变动对婴儿死亡漏报率更敏感；死亡水平及婴儿死亡漏报可能性指数的分布全局上不存在显著的双变量自相关，但局部显示一定的空间聚集且分性别的空间模式具有相似性。

关键词: 婴儿死亡漏报, 平均预期寿命, 婴儿死亡漏报可能性指数, 聚类分析, 空间自相关