| Abstract [eng] |
One of the goals is to create working models that would predict the costs of income under the Social Insurance Fund, separate groups of pension costs in accordance with the number of maternity / parental benefits under the Social Insurance Fund and the number of insured earnings in accordance with the following variables: compulsory insurance policyholders' premiums, the insured payments, self-employed contributions, allocations from the Ministry of the state budget, other sources of income. Models set up by examining what relationship existed between the variables. Of identical, or can predict values for the selected variables, linear regression analysis or multiple regression analysis. Further work is to predict old-age pensions in Lithuania according to two systems: points and virtual. In order to make forecasts based on a points system, is transformed by the current pension calculation formula. Estimate of pensions under a virtual system, whichever is the formula used in Germany. Retirement pension forecast made in accordance with these systems need to predict the insurable earnings, average earnings of the insured persons the average wage, the number of pensioners, the number of insured persons and the number of unemployed. The comparison points to a virtual old-age pension systems, it was observed that the points system is useful for individuals who receive lower than average income and insured with the necessary experience. Individuals who experience greater than that needed, or receiving more than the average insured income benefits for a virtual system. |
