| Research BackgroundThe mortality level of Chinese is declining,the average life expectancy is increasing,and the risk of longevity is prominent.Moreover,regional development is uneven,and there are great differences in the mortality patterns of the population in different provinces,and the study of mortality rates by province can more comprehensively and accurately grasp the death patterns of Chinese and measure the regional differences in mortality.The formulation of government public service policies will be directly affected by the longevity level of residents,such as the income and distribution of social pensions.In addition,the mortality level of each region will also directly affect the business decisions of insurance companies,such as the pricing of life insurance products and the evaluation of reserves.Based on the above analysis,it is of great practical significance to explore the estimation and dynamic evolution of mortality in different provinces.Research ContentThe research content of this paper is mainly divided into three parts:first,based on the mortality data of the three censuses in 2000,2010 and 2020,through the Lee-Carter model under limited data,and through the"double stochastic process with drift term",the mortality rate of men and women by age at the national level and in various regions is estimated by singular value decomposition method(SVD),and the mortality rate predictions of various regions in China from 2000 to 2020 are obtained.The estimated results are analyzed.Second,based on the estimation results of population mortality in each region,the standard deviation plot and sequence plot of logarithmiclog(m_x),mortality age factor a_x and b_x estimates were constructed,and the differences in mortality in each region were analyzed.Third,the population mortality rate in each region in the next 15 years(2021-2035)is predicted,and the future dynamic evolution of population mortality in Chinese provinces is analyzed based on the mortality forecast results at the national level and provinces from 2021 to 2035.Research ConclusionsFirst,from the census data.In terms of the time dimension,China’s population mortality rate has shown a gradual decline over time;in terms of the age dimension,there are differences in the level of mortality improvement among different age groups,with the 0-year-old group showing the most significant improvement and the older population showing a lower level of improvement;in terms of the gender dimension,the level of mortality for females is generally lower than that for males,but the difference between male and female mortality rates has gradually decreased over time.Second,the results from the population mortality estimates.From an age perspective,the sensitivity of male and female age groups to changes in mortality declines continuously with age,being highest in infancy and lowest in the older age groups;from a gender perspective,the mortality dynamicsk_tfor males show a steady decline over time,indicating that male mortality tends to decline steadily,while k_t for females shows a fluctuating decline over time.Third,the results are seen in terms of regional differences.In the time dimension,there are differences in mortality levels across regions,but the differences decrease gradually over time;in the age perspective,the differences in mortality levels across regions are greatest in the 0 and 100+age groups and least in the 80-84 age group;in the gender perspective,the differences in mortality rates across regions are higher for females than for males,and the mobility of young and middle-aged people causes an increase in geographical differences in mortality rates for 15-49 year olds,especially for males.Fourth,the evolution of population mortality dynamics is seen.Temporally,mortality levels will continue to decline in the future in all regions,but there is no significant change in the level of mortality improvement in the 60-year age group;based on a gender perspective,mortality rates for males in the 60-year age group are higher than for females overall over the next 15 years,but the geographical differences in mortality rates for males are lower than for females overall.Research ContributionsFirst,the content aspect.This paper enriches research related to mortality in China’s sub-provinces by analyzing population mortality rates and geographical differences and dynamic evolution from an inter-provincial micro perspective.The results of this paper will provide practical references and recommendations for the pricing of life insurance products,the formulation of pension insurance policies and the estimation of the demand for social public services.Second,the methodological aspect.Using cross-sectional and time-series information on population mortality,this paper is the first to apply the Lee-Carter model with limited data to sub-provincial population mortality projections.In the existing literature,Yang(2019)only used data from the sixth census in 2010 in his study of sub-provincial population mortality,using only cross-sectional information on mortality without considering information on time series,and did not use a model to make projections of future population mortality.Third,the data aspect.The problem of poor quality and low data on population mortality is addressed.This paper addresses the issue of poor data quality by using three higher quality census data for the 2000,2010 and 2020 censuses for the analysis.The paper uses the Lee-Carter model with limited data,which allows modelling of limited population mortality data and hence extrapolation of mortality rates for the projection years,addressing the issue of less data. |
