Abstract: This paper analyzes death underreporting in China＇s population censuses with a reestimation of levels and patterns of mortality since 1990．There is severe death underreporting at ages 0-4，more notably in the 2010 census．Death reporting is also rather incomplete at ages 60 and over． Death probabilities are reestimated by age and sex from which life tables are reconstructed． Over the last 20 years，life expectancy at birth has increased from 67.33 to 71.58 years for males and from 71.70 to 78.26 years for females，respectively．Pace of mortality improvement has been more rapid in the recent decade that in the past．

Keywords: Death Underreporting, Cohort Survival Method, Brass Logic Life Table Method, Age Patternof Mortality

摘要： 文章采用留存率方法和布拉斯罗吉特生命表法对1990年以来三次人口普查的低龄人口和老年人口的死亡漏报、年龄别死亡概率、寿命水平进行重新评估计算。结果发现,低龄人口0～4岁漏报较为严重,特别是2010年第六次人口普查,不仅仅0～4岁人口死亡漏报严重,而且60岁及以上老年人口的死亡漏报也相当严重。根据重新修正的分性别、分年龄别的死亡概率数据,构建生命表,分别计算得到1990年、2000年和2010年出生预期寿命,1990年我国男性出生预期寿命为67.33岁,女性为71.70岁,2010年男性为71.58岁,女性为78.26岁。进入21世纪的前10年我国寿命水平提高快于2000年之前的20年。

关键词: 死亡漏报, 队列留存法, 布拉斯罗吉特生命表, 死亡模式