A Study Of Mortality Differences In China Based On Cluster Analysis

According to the major data communique of the national 1% population sampling survey in 2015,the total population of 31 provinces(autonomous regions and municipalities)in Chinese mainland in 2015 was 1.373 billion(including active servicemen).According to the World Population Prospects(2017)released by the United Nations Population Division in July 2017,China's population accounts for 18.9%of the world's total population,which is 1.9 times that of Europe.The mortality rate level of a country or region is not only an important indicator to measure the natural changes of the population,but also an objective standard to measure and evaluate the health status,welfare level,socio-economic and health care progress of the residents.The social living conditions,especially the social and economic conditions,will have different effects on death.China has a vast territory and a large population.At the same time,it is also a big family of 56 ethnic groups.There are great differences in medical and health service conditions and economic and cultural development level among provinces,cities and districts in China,which will inevitably lead to discrepancies in population mortality of different regions.The differences of population mortality across provinces(autonomous regions and municipalities directly under the central government)(hereinafter referred to as "different provinces")and the underlying reasons are worth pondering.The specific mortality rate and life span of each province have an important impact on the demand estimation and policy parameters of public services provided by local governments,such as the overall social security policy.Therefore,the research on the mortality rate among provinces and cities in China has certain practical significance.In addition,by studying the mortality differences and their changes in different regions of China,it is helpful for life insurance companies to price their products accurately,and to assess the future longevity risk more accurately,so as to better maintain the company's sustainable and stable operation.In previous studies,although some researchers have compiled the life tables of provinces and predicted the life expectancy,they have not considered the impact of geographical regions,economic environment and other factors on neighboring provinces,nor further quantified the impact of mortality differences on the price of life insurance products,and put forward suggestions on the product design of insurance companies on this basis.The data source used in this thesis is the sixth census data from the National Bureau of statistics.Whittaker smoothing method is used to obtain the specific mortality data of each province.In addition,considering the nature of the data,using unsupervised learning hierarchical clustering,taking the European distance as the classification standard,the provinces in China are divided into six regions according to the similarity of mortality,and the geographical differences of different groups of provinces are analyzed.Finally,based on the grouped mortality data,this study calculates the premium differences of term life insurance products,life insurance products and annuity insurance products in different regions,and puts forward some suggestions for the future product development of insurance companies.The results have shown that the similarity of mortality is not only affected by geographical location,but also by the economic situation of each province.The difference of mortality rate among different age groups will be more obvious in provinces with far different geographical location and unbalanced economic development.This provides a new idea for insurance companies to price for people from different regions,such as adopting more stringent underwriting standards or increasing fees to a certain extent for people in provinces with poor mortality rate;or,provincial insurance companies can develop their own insurance to meet the specific needs of the local population.In the development of China's insurance industry,the homogenization of insurance products has remained a problem.The reasons include not only the strong supervision of the insurance industry,but also the investment orientation of shareholders.Up to now,the number of life insurance companies in China has reached nearly 100,but the degree of product homogeneity is still high.With the development of the Internet,more and more companies begin to use big data and other new technologies to develop products,speeding up the pace of product differentiation.However,as the main sales channels of the entire insurance industry are still offline,the market share of Internet online channels is relatively low,so its influence on the entire industry is limited.In the future,insurance companies should get rid of the low-quality competition,change their own development ideas,no longer take the sales and scale effect as the guidance,but take the real needs of consumers as the guidance,constantly improve their product connotation and after-sales service,and design more differentiated products,so as to truly achieve "insurance" ——providing sufficient and reliable guarantee for the people.The mortality rate varies greatly among different regions,which is both a risk and an opportunity for insurance companies.On the one hand,when designing and pricing products,insurance companies can understand the main sales areas of their products through empirical analysis,and set different rate assumptions to fit the actual structure of the insured,so as to avoid the occurrence of over indemnity risk to a certain extent;On the other hand,insurance companies can design insurance with different responsibilities and claims conditions for different people in different regions to carry out differentiated marketing,so as to occupy the market segments and expand the scope of business.