Research On Prediction Model Of Aging Coefficient In Zhejiang Province

With the decreasing in total fertility rate and increasing in average life expectancy, the process of population aging in Zhejiang Province has been accelerating since reform and opening. According to statistical data, the aging coefficient in Zhejiang Province was only 4.9% in 1978, and it has already reached 7.07% by the end of 1988. Moreover, it has reached 11.3% by the year of 2008. Zhejiang is one of the provinces with serious population aging problem in China. The worse of population aging not only aggravates the society burden, but also undermines the rational allocation of the human resources. Therefore, it will play an important role in social security, economic development and other aspects of Zhejiang to understand the development process of aging coefficient and the causes correctly in Zhejiang Province, and will be also important to predict the trend of aging coefficient and propose effective measures.Based on the household population data in Zhejiang Province from 1978 to 2008, I build vector autoregression models to predict total population, aging population and aging coefficient from 2009 to 2050.The methods of building models are data mining, demographic, and the theory of demographic. Then I propose some suggestions. The idea of paper is as following:Firstly, I introduce the history of total population, aging population and aging coefficient in Zhejiang Province. And analyze the causes of aging in Zhejiang Province.Secondly, I propose a theorem about the relationship between average life expectancy and the number of surviving. After proving the theorem, I predict complete life tables in 2010, 2020, 2030 and 2040. Using four life tables, I estimate the net increasing of aging population from 2011 to 2050. And propose a proposition by which we can determine the long-term changes trend of aging population.Thirdly, with the household population data in Zhejiang Province from 1978 to 2008, I make granger causality tests, ADF unit root test and Johnson cointegration tests. From the tests I find out the main factors that influence total population or aging population. Then qualitative and quantitative analysis are combined to establish the vector autoregression models; I design four different programs about average life expectancy, population of provincial net intake and total fertility rate, then predict dynamically about the total population and aging population from 2009 to 2050.Finally, I calculate the aging coefficient by the predicted total population and aging population, and analyze the trends of them. With predictions I propose some suggestions to help the government respond to population aging.