| The pension insurance institution is the most important part in the society security institution, and it is formed with the change of society which is bought by the revolution of industry and the demand of society which is bought by the development of economy, which itself doesn't create wealth and only make individual income maximum on the time travel compliably or uncompliably, so that it can in effect eliminate the risk because of descending of the capability of earning when people were old, and enable every economic entity in the system to be supported when he was old.With the development of society, the average person's life and the time of living after retiring is longer, the global demographic ageing is in evidence, the demographic support ratio is raising, the pension insurance becomes more important, and the challenge of the pension insurance institution is also more and more rigorous. Compared with the west developed country which put the pension insurance institution in practice earliest, the problem of us is more rigorous and more complex.The conception of microsimulation model is first put forward by professor Orcutt of Yale in America in 1957, which was used as the method of constituting and modifying the policy of society and economy in the decision-making department of government. At present, it is building harmonious socialism in our country, the various measures of reform, such as the reforms in tax, the society security, medical treatment and housing, need to be perfected and actualized. The pension insurance institution is one of the most important part in the society security institution of China, but the income and expenses of pension is being out of balance, the pressure of finance expend is more and more strength. Microsimulation model will help to analyze and estimate the reform projects and measures in our country without question. Therefore, it is necessary to use foreign technology and experience for reference and accelerate to develop and apply this model in China.Microsimulation is a process that takes individual, family or corporation as the object which it describes and deals with, and simulates real economic system by computer. In the model of microsimulation, macro-economy is composed of a great deal of micro-unit, the situation of function of macro-economy and the economic policy will have profound effect on the situation of micro-unit, and the changes of the situation of micro-unit roll up the changes of the situation of the situation of macro-economy. Because our micro-data are often discrepant with the economic environment which we want to analyze, so many times , we need to translate the initial data into the data which can reflect the real economy by aging. It has two kinds of aging: static aging and dynamic aging (corresponding to static model and dynamic model). Static aging translates the initial period data into target period by one step, often takes the method of reweighting and adjusting attributes. Dynamic aging translates the initial period data into target period step by step. Static model is used to analyze the immediate effect of economic policy, such as tax system and living safeguard system at the lowest, while dynamic model is used to analyze long period effect of economic policy, such as financial effect and income distributional effect of pension insurance system and unemployment insurance system.In the process of setting up a microsimulation modeling, one of the main difficulties of the work is to determine the probability of the occurrence of an event distribution. The probability distribution of the incident can be divided into theoretical distribution and experienced distribution. In the study of social economic problems, it frequently used experienced distribution models, but the distributions computed only from one group of observational data have greater randomness, and in the less data range, error is relatively greater, and did not exceed the maximum and the minimum scope of the distributions. In the life events (including death, education, marriage, divorce, birth, etc.), the error of the model mainly come from these parameters in the life events, namely the deviation of the probability distribution of the change of the characteristic variables. Therefore, determining model parameters is a very important and difficult step in the process of setting up a microsimulation model.This paper developed a dynamic microsimulation model used to analyze the financial effect and the income distributional effect of the reform of the pension insurance system in our country. The simulation method of the model is based on the incident, and as the time goes forward, the model evolutes. The process and the account of pension savings is closely related to sex, occupation, industry, income and work experience, and these factors is related to the major life events of the population, such as birth, education, marriage, employment, retirement. Therefore, each individual is described by demographics and economics information, including age, income (wages, pensions, etc.), relations, employment in the labor market, such as occupational and industry. Each individual must experience two major categories of events: (1) life events, including the death, birth, marriage, employment, etc. (2) political events, including raising pensions and pension payment. The population in the micro-data is aging from 2006 to 2045 year by year, according to the probability, each event randomly designated to the appropriable individuals, and the status of the individuals are revised according with the different events, the accumulation of the change of the status educes the macroeconomic variables.In particular, this paper has made some improvements in determining the parameters of the model, using Logistic model to analyze the determinants of life events, and estimate the probability of the incidents under certain conditions, with a view to gaining more accurate parameters. Logistic model is used in the circumstances when linear regression model is limited. For example, when the dependent variable is a classification variable rather than a continuous variable, linear regression model does not apply. The commonly used statistical method is a log-linear model, when the dependent variables in a log-linear model is a Category 2 or more variables and is defined as a function of a number of variables, the log-linear model will become Logistic regression model. Based on the 2005 survey data in Jilin Province, this paper estimates parameters in the modules using Logistic model. In the birth module, according to the age and level of education of the married women of childbearing age but not childbearing yet, we estimate the probability of reproductive first child. In the education module, according to the cultural level of the micro individual's parents and the average left school age of the group which is the same age with the micro individual, it is estimated that the possibility of the micro individual's highest level of education in elementary school, junior high school, high school, college, undergraduate and postgraduate. In the marriage module, whichever is male, according to the cultural level of men and the age difference with their spouses, it is estimated that the possibility of the level of education of their spouses. In the employment module, in accordance with the individual's level of education and experience, we distribute industries for the new employee, as also considering their gender, distribute vocations for the individual, for the individual in the conditions of the promotion, taking their level of education experience and gender into consideration, determine the possibility of promotion, and adjust the corresponding attribute. In the income module, for the stabilized employment staff, we assume that their annual growth rate of wage is 7%, for new employment individuals, according to their industries and occupations, we evaluate their income, and for the individuals with promotion events, according to the average wage of their industries and occupations, educational level and working experience, we assign their income.Finally, the microsimulation model update the individual state by a series of life events and policy events, and accumulated by the natural state of the individual educed the relevant regional economic variables. Comparative analysis of the regional economic variables before and after the adjustment of the pension insurance system, we can see that the implementation of the new policy would expanse the income gap in society, but can narrow the gap scale of the pension insurance fund, smooth the financial subsidies, stabilize the pension replacement rate, increase the rate of personal pension insurance yield. Thus, the reform of pension insurance system basically achieves the goal of balance the fairness and efficiency in the public policy. |
PDF Full Text Request