摘要
充分利用男女性别人口死亡率数据的有效信息,联合预测男女人口死亡率,能够提升死亡率建模方法的科学性和预测结果的合理性。选取国家统计局公布的分年龄、分性别人口死亡率数据,采用多人口Li-Lee模型对男女性别人口死亡率进行联合预测,并将预测结果与单人口Lee-Carter模型进行比较,以探寻适合中国人口死亡率预测的模型方法。研究表明,相对于Lee-Carter模型,Li-Lee模型在中国男女性别人口死亡率拟合与预测中具有较好的表现,全样本拟合值的绝对百分比误差平均降低了0.23%,且短期预测值精确度相对较高。同时,Li-Lee模型预测的人口死亡率性别比在长期中能够进入合理区间,避免了出现Lee-Carter模型预测结果中的异常数值,并通过补充中国台湾省人口死亡率数据,进一步验证了Li-Lee模型长期预测结果的稳健性。研究结论能够为我国积极应对人口老龄化、落实人口中长期均衡发展战略提供基础性支撑信息。
Making full use of the effective information of male and female mortality data and jointly forecasting male and female mortality can improve the scientificity of mortality model and the rationality of forecasting results.This paper selects the age and sex specific mortality data published by the National Bureau of statistics of China,uses the multi-population Li-Lee model to jointly model the mortality by both sexes,and compares the forecasting results with the single population Lee-Carter model,so as to explore the model method suitable for the prediction of China’s population mortality.The result shows that Li-Lee model is better than Lee-Carter model in fitting and forecasting the mortality of both sexes in China.Further,the MAPE value of full sample fitting performance obtained by Li-Lee model is reduced by 0.23%on average,and the short-term forecasting value is more accurate.At the same time,the mortality ratio predicted by Li-Lee model can enter a reasonable range in the long run,avoiding the abnormal values of Lee-Carter model in some age.Furthermore,by supplementing the mortality data of Chinese Taiwan,this paper further verify the robustness of the Li-Lee model.The research conclusion can provide basic information support for China to actively deal with population aging and implement the medium and long-term development strategy of population.
作者
赵明
ZHAO Ming(School of Finance,Capital University of Economics and Business,Beijing 100070,China)
出处
《人口与发展》
CSSCI
北大核心
2023年第2期61-72,共12页
Population and Development
基金
全国统计科学研究重点项目“新时代人口长期均衡发展综合评价指标体系研究”(2022LZ36)
首都经济贸易大学青年学术创新团队项目“大数据背景下健康老龄化综合评价指数体系与精算测度研究”(QNTD202104)
北京市属高等学校优秀青年人才培育计划项目“大数据背景下老年人健康预期寿命不平等的多维测度研究”(BPHR202203166)
北京市教育委员会社科计划一般项目“积极老龄化视角下北京市低龄老年人口劳动参与意愿与影响机制研究”(SM202110038005)~~。
作者简介
赵明(1984-),男,首都经济贸易大学金融学院,副教授,博士生导师,研究方向:人口死亡率建模与长寿风险量化管理。
