利用DCMD模型生命表系统对“六普”数据中死亡漏报的估计

The Estimation of Death Underreporting in the 2010 Population Census Based on DCMD Model Life Tables

低龄死亡概率和老龄死亡概率对估计0岁和60岁预期寿命有特殊意义,直接关系到计算结果的准确性,且涉及养老保险等一系列政策的制定。众多研究表明"六普"数据中低龄组与老龄组死亡数据存在明显漏报,并且给出了不同的修正方法以及相应的结果。文章利用DCMD模型生命表系统,根据"六普"数据中60~75岁的分性别人口数计算出老龄死亡率,结合UNICEF的儿童死亡率和IHME的成人死亡率数据,对"六普"时期我国分性别的国民生命表进行重建,重新计算了0岁和60岁的预期寿命,并在此基础上估计了低龄组和老龄组死亡数据的漏报程度。结果表明,"六普"时期我国婴儿死亡率男性为16.41‰、女性为15.94‰,相应的漏报水平为77.3%和75.5%;老龄死亡概率(60~89岁)男性0.703、女性0.595,相应的漏报水平为2.3%和7.0%。

Young-age and old-age death probabilities are essential for accurate estimation of life expectancy at age 0 and 60, which have maior implications for the formulation of endowment insurance and other policies. Previous studies have testified that death underreporting commonly exists in the mortality data of young-age and old-age group,such that various approaches of modification were offered. This paper recalculated sex-specific life tables of China and life expectancy at age 0 and 60 for the 2010 population census based on the DCMD model life tables, using child mortality data from UNICEF, adult mortality information from IHME, and old-age mortality rates derived from the sex-specific population aged 60 - 75 in the census.The unclerreporting rates of young-age and old-age mortality are estimated. The result shows that in the 2010 population census adiusted infant mortality rates are 16.41% for male and 15.94% for female,implying an unclerreporting rate of 77.3% and 75.5% respectively;the adiusted old-age （from 60 to 89） death probabilities are 0.703 for male and 0.595 for female with the underreporting rate being 2.3% and 7.0% respectively.