The Measurement Of China's GDP Growth Rate

China's GDP accounting is in a process of continuous improvement,now the accounting is divided into the national level and regional level accounting.Among them,the national annual GDP data includes the production accounting and expenditure accounting data.Due to the difference of statistical data source,the characteristics of the accounting and inevitable statistical error,the countries and regions,and between two kinds of accounting methods of GDP data exist differences in values.This paper summarizes the development of China's system of national accounts and project structure characteristics of production method and expenditure method of China's current price of GDP.Based on which,this paper selects consumer price index and price index of investment in fixed assets and so on to process the price factors,according to the weight of each part to calculate the real GDP growth rate in expenditure method.By selecting the growth of urban employed population,the various indicators of broad money supply to regress the real GDP growth in production method and expenditure method,we find that the information contained in each has its strong point.Regarding the GDP growth rate of production and expenditure method as the real GDP growth forecast,through the method of forecast combination discussed under the condition of the quadratic loss function how to obtain the optimal combination.Also this paper analyzes,in detail,the advantages of this method in reducing the estimation error,and explains the economic cycle reflected by the estimated GDP growth at the condition of the current domestic and foreign economic.Assumed that regional and national level production method GDP growth rate are both the measured values of the real GDP growth rate and real GDP growth rate is AR(1)process,based on the Kalman filter state space model,this paper estimates the real GDP growth rate,and compares the 95% of the smoothing confidence interval by the process of the iteration with the growth rate results calculated by the forecast combination model.