Study on the trend of mortality rate and measurement of longevity risk in China
(2.Chinese Academy of Labor and Social Security)
(3.School of Statistics, Renmin University of China)
【Abstract】Using the statistics of mortality rate from 1950 to 2018, this paper studies the trend of mortality rate in China, and establishes a stochastic mortality model to measure the longevity risk faced by China’s pension system. The results are as follows. (1) China’s crude mortality rate declined rapidly from 1950 to 1981. The pace of decline increased first, then decreased with age, and there is little difference in gender. (2) China’s crude mortality rate declined steadily from 1981 to 2005. The rate of reduction in infant mortality has increased greatly, and that for the males aged 15–59 has slowed down significantly, which has resulted in a lagged-behind life expectancy for males as compared with females. (3)From 2005 to 2015, the crude mortality rate increased, which is resulted from the changes in population age structure. Meanwhile, the age-specific mortality rate still maintained a downward trend. The mortality rate of the population over 60 years old declined faster than before. (4) In population mortality prediction, the robust model takes the trend of mortality change into consideration, and hence enhances the validity of prediction. (5) The robust model has lower modeled risk in measuring longevity risk and is more conservative in longevity risk estimation. Results from this model can retain more risk reserve for the pension system.
【Keywords】 population mortality; trend; longevity risk; measurement model;
(Translated by DUAN yinhong)
. ① Since the age effect factor uses the concept of change from the last period in statistics, a geometric average method should be used to characterize the average decline in mortality over a period of time. This indicator is used by the North American Society of Actuaries, which is applicable to data with fewer years can be obtained. The average annual change rate of mortality over a certain period of time can be calculated, and the comparability is strong. [^Back]
. ② The United Nations World Population Division publishes World Population Prospects: The 2017 Revisionevery two years, and uses Bayesian stratified model methods to estimate the age-specific mortality data of countries in the world. The method is scientific, the results are reasonable, and it has certain authority in the world. It has strong reference value for this study. [^Back]
. ① The population mortality rate generally shows a downward trend with time, so In [m (x, t + 1) / m (x, t)] is a negative value. After adding the minus sign, it is used as an approximation of the relative level of mortality reduction. [^Back]
. ② The Chinese male population mortality rate is selected as the representative of the study because the population change sampling is used in most years, and the female population mortality rate is significantly lower than that of males. For some ages, females have zero death, which leads to distortion of mortality rates. It is less common for male population. [^Back]
. ① In 2015, the infant mortality rates in the United States, Sweden, and Japan were 6.394‰, 2.785‰, and 2.080‰, respectively; in 1985, the average age fractions of the adult population in the United States, Sweden, and Japan were 5.108‰, 3.621‰, and 3.306‰. [^Back]
. ① For the mortality prediction of the population aged 100 years old or above, extrapolation can performed by the mortality model of the elderly population, such as Gompertz model. [^Back]
1. Poston, D. & Fang, D. Population Research (人口研究), (3) (1991)
2. Huang, K. Population Research (人口研究), (5) (2015)
3. Li, H. et al. Chinese Journal of Health Statistics (中国卫生统计), (3) (2015)
4. Wang, X. & Mi, H. Statistical Research (统计研究), (2) (2013)
5. Wang, X. & Zhao, M. Statistical Research (统计研究), (9) (2014)
6. You, Y. Population & Economics (人口与经济), (6) (1984)
7. Zhai, Z. & Tao, T. Chinese Journal of Population Science (中国人口科学), (1) (2010)
8. Zhao, M. & Yang, F. Population Research (人口研究), (5) (2013)
9. Zhao, M. & Wang, X. Statistical Research (统计研究), (12) (2015)
10. Zhou, Y. et al. Chinese Journal of Population Science (中国人口科学), (3) (1989)
11. Bongaarts J. (2005), Long-range Trends in Adult Mortality: Models and Projection Methods. Demography. 42 (1): 23–49
12. Booth H. (2006), Demographic Forecasting: 1980 to 2005 in Review. International Journal of Forecasting. 22 (3): 547–581.
13. Bulatao R. A., Bos E., Stephens P. W., Vu M. T. (1989), Projecting Mortality for all Countries. WPS 337. Population and Human Resources Department. The World Bank.
14. Chu C. Y. C., Chien H. K., Lee R. D. (2008), Explaining the Optimality of U-shaped Age-specific Mortality. Theoretical Population Biology. 73 2): 171–180.
15. Girosi F., King G. (2008), Demographic Forecasting. Princeton: Princeton University Press.
16. Hamilton W. D. (1966), The Moulding of Senescence by Natural Selection. Journal of Theoretical Biology. 12 (1): 12–45.
17. Horiuchi S., Wilmoth J. R. (1998), Deceleration in the Age Pattern of Mortality at Older Ages. Demography. 35 (4): 391–412.
18. Holzmann R., Palmer E. (eds) (2006), Pension Reform:Issues and Prospects for Non-financial Defined Contribution(NDC)Schemes. Washington D. C. The World Bank Publications.
19. Lee R. D., Carter L. R. (1992), Modeling and Forecasting U. S. Mortality. Journal of the American Statistical Association. 87 (419): 659–671.
20. Li N., Gerland P. (2011), Modifying the Lee-Carter Method to Project Mortality Changes up to 2100. Paper Presented at the Meeting of the Population Association of America.
21. Li J. S., Hardy MR., Tan K. S. (2010) Developing Mortality Improvement Formulas: The Canadian Insured Lives Case Study. North American Actuarial Journal. 14 (4): 381–399.
22. Li N., Lee R, Gerland P. (2013), Extending the Lee-Carter Method to Model the Rotation of Age Patterns of Mortality Decline for Long-term Projections. Demography. 50 (6): 2037–2051.
23. OECD(2011), Pensions at a Glance: Pensionable Age and Life Expectancy, 1950–2050. In Pensions at a Glance2011:Retirement-income System in OECD and G20 Countries. Paris: OECD Publishing.
24. Plat R. (2011), One-year Value-at-Risk for Longevity and Mortality. Insurance:Mathematics and Economics. 49 (3): 462–470.
25. Purushotham M, Valdez E., Wu H. (2011), Global Mortality Improvement Experience and Projection Techniques. Society of Actuaries.