中国死亡率的抽样误差修正、预测及应用——基于引力模型与日本相应数据

Sampling Error Correction,Forecast and Application of China Mortality——Based on Gravity Model and Corresponding Data in Japan

死亡率引力模型是研究两个相关但暴露数规模不同的人群的动态死亡率模型。本文针对中国人口死亡率抽样数据偏差较大这一特点,运用引力模型思想借鉴日本相应数据建模。研究发现,引力模型改进了我国人口未来死亡率建模的方法,提供了更准确的预测。对于中国现有的数据,如果不考虑引力效应会导致年金产品的过低定价,低估未来老龄人口的抚养比。

The gravity model is used for studying the mortality rate dynamics between two related but different-sized population groups. In line with the large sample data bias of population mortality in China, we applied the gravity method to model it, together with a reference to the corresponding data in Japan. We found that the gravity model improved the modeling method for the future mortality of our country, and provided a more accurate forecast for the future mortality rate of China＇ s population. For China＇ s existing data, the neglect of gravitational effects would lead to the underpricing of annuity products and underestimating the future dependency ratio of the aging population.