Comprehensive Evaluation Of Fiscal Expenditure Based On The Principal Component Regression

Since the reform and opening-up in China, especially after establish the marketeconomic system, the public financial expenditure structure has been changed makes thereform of China’s financial system achieved remarkable success, The focus on the fiscalsystem reform is reform the fiscal allocation relationship, including the reform of thedistribution relationship between central and local government. Reform of local finance isboth an important goal of China’s financial system and our socialist market economic reform.Therefore, it is particularly important to adjust and optimize the structure of fiscal expenditurestructure. adjust and optimize the fiscal expenditure structure is will be an important task tothe public financial reform and development facing in a certain period. And beginning withthe research of the historical evolution of our country’s public financial expenditure, analyzesproblems exist in the expenditure structure.summarize and put forward a proactive fiscalpolicy, through comprehensive and detailed analysis of the main large and medium-sizedcities of china and analysis the comprehensive guidance factor affecting the development ofregion, will helps to provide theoretical guidance and recommendations for the sustainabledevelopment of the city’s economy.This article mainly aims at the Eighth part of “China Statistical Yearbook2011":which isthe local fiscal spending. The total number of our country’s financial Expenditure ofprovinces and autonomous regions(not including Hong kong, Macao, Taiwan area specificallydesignated in the state plan is31).In view of the31City, it lists13economic indexes. Thispaper’s aim is studying the economic indexes of the31cities.This article uses the experimental method is regression analysis、principal componentanalysis、principal component clustering and co-integration analysis.(1) Regression analysis is a statistical method to identify directional relations of thevariable dependences which have extensive application value. According to the relationshipbetween the explanatory variables and the explanatory variables types, regression analysis canbe divided into linear regression analysis and nonlinear regression analysis;(2) Principal component analysis is a statistical analysis method which use multiplevariables into a comprehensive less important variables. (3) The ideas of principal component clustering is take main component then takeseveral component do cluster analysis, calculate component comprehensive score, rankingand evaluate of the urban development.(4) Co-integration analysis. time-series variables have co-integration relationship and cando co-integration analysis which must meet time-series variables are stable and balancedrelationship.