Population Research ›› 2015, Vol. 39 ›› Issue (5): 37-48.

Previous Articles     Next Articles

The Application of Lee Carter Model to the Extension of Model Life Tables: The Case of China Regional Model Life Tables

Huang Kuangshi   

  1. China Population and Development Research Center, Beijing 100871
  • Online:2015-09-29 Published:2015-12-11
  • About author:Assistant Professor, China Population and Development Research Center

Lee Carter模型在模型生命表拓展中的应用——以中国区域模型生命表为例

黄匡时   

  1. 中国人口与发展研究中心,北京 100871
  • 作者简介:中国人口与发展研究中心助理研究员
  • 基金资助:

    国家社会科学基金(批准号:14CRK005)“基于LEE CARTER模型的中国模型生命表拓展研究”

Abstract: The classic LeeCarter model decomposes the change of mortality into three parts: the age-specific average mortality rate, the age-specific change of mortality rate, and the grand tempo change of mortality rate. Such decomposition not only considers the different baselines of mortality for various populations, but also specifies the two types of mortality changes. Because we can obtain different age-specific mortality levels by adjusting either the age-specific change rate or the grand tempo change rate, the Lee Carter model can be used to extend the model life table. However, the classic Lee Carter model would amplify the gender gap of life expectancy and also result in unreasonable mortality ratios among different ages in the long-term mortality forecast. This study modifies the classic Lee Carter model with coherent model and rotation model to overcome its deficiencies in the long-term mortality forecast. The modified Lee Carter model can be used to both project the age-specific mortality rate at a high level of life expectancy and extend the model life table. We take China regional model life table as an example to show the application of the model life table extension with the modified Lee Carter model.

Keywords: Lee Carter Model, Model Life Table, China Model Life Table, Mortality Forecast

摘要: 经典Lee Carter模型将死亡率变化分解为年龄别平均死亡率、年龄别变化速度和时间变化三部分,既考虑了不同人群的死亡率基数,也细分了死亡率未来变化的两个重点,通过调整年龄别的变化速度和时间变化可以生成不同预期寿命水平的年龄别死亡率,在模型生命表由较高水平的预期寿命向更高水平的预期寿命拓展上具有独特优势。不过,该模型在死亡率长期预测上有性别差异扩大和年龄别比值失调等问题。文章对该模型进行了相干改进和倒置改进,修正了该模型在死亡率长期预测上的不足。改进的Lee Carter模型既可用来预测高预期寿命水平下的年龄别死亡率,也可用来对模型生命表拓展。文章以中国区域模型生命表为例介绍了改进的LeeCarter模型在中国模型生命表拓展中的应用。

关键词: Lee Carter模型, 模型生命表, 死亡率预测, 预期寿命