Abstract: Continuous and complete population data by age and education level are the basis for studying the dynamic relationship between population and economic and social development in a country or region. However, in non-census years, such data cannot be directly obtained or its quality cannot be guaranteed. To this end, the national population census data from 2000, 2010, and 2020 are used to reconstruct the population data by age and education level for non-census years from 2000 to 2020 using the population discrete development equation forward and backward bidirectional feedback fitting method. Based on this, an improved accounting model is used to calculate the human capital stock from 2000 to 2020.The research results show that the difference between the reconstructed non-census year total population data and the adjusted data from the National Bureau of Statistics is very small. The reconstructed age-specific population data corrected the bias in the age composition data published in previous statistical yearbooks. The calculated data on human capital stock shows that it increased from 15.856 billion person-years in 2000 to 26.479 billion person-years in 2020, with an average annual growth rate of 2.60%. Affected by the ageing population and the decrease in the working age population, the growth rate of human capital stock shows a decreasing trend between the first and the second ten-year periods.

Keywords: Population Census, Age Structure, Educational Attainment, Human Capital

摘要： 连续、完整的分年龄、分受教育程度人口数据是研究一个国家或地区人口与经济社会发展动态关系的基础。然而在非普查年份,这类数据要么无法直接获取,要么质量得不到保障。为此,利用2000年、2010年和2020年全国人口普查数据,采用马尔科夫链-人口离散发展方程正反双向反馈拟合方法,重构出2000~2020年期间非普查年份的分年龄、分受教育程度人口数据。在此基础上,利用改进后的核算模型计算出2000~2020年的人力资本存量。研究结果显示,重构的非普查年份总人口数据与国家统计局调整后的数据差距很小,重构的分年龄人口数据修正了国家统计局公布的人口年龄结构数据存在的偏差;中国人力资本存量从2000年的158.56亿人年攀升到2020年的264.79亿人年,年均增长率为 2.60%;受人口年龄结构老化和劳动年龄人口减少的影响,人力资本存量增长速度呈现放缓趋势。

关键词: 人口普查, 年龄结构, 受教育程度, 人力资本