Models, Methodologies And Empirical Research On Measuring Tax Capacity

Tax Capacity refers to the maximum potential tax revenue reserves under the tax systemestablished in a country or region. Measuring tax capacity has been of great concern. Basedon the literature review of related research, and integrated using of mathematical economicanalysis, econometric modeling of panel data, stochastic frontier analysis techniques, datadistribution estimation and stochastic simulation techniques and so on, this dissertationstudies the related concerns and theories associated with tax capacity, development ofChinese financial and taxation system, models, methodologies and demonstration onregional tax effort indices and (absolute) tax capacity of sub-types tax. And the conclusionsand innovations of this research are as follows:Firstly, the dissertation defines the related concepts of tax capacity and their relationships,studies the Chinese tax-sharing fiscal system and tax policy, practices of monitoring taxsources and environmental status of tax data, reveals that the formation andimplementation process of tax capacity in a country or region is the essential feature of thetax production system and puts forward the hypodissertation of the systematic taxproduction about the formation and implementation process of tax revenue (or capacity).At the same time, it was found that with the further reform and opening up to the world,the potential capacity of tax revenue in every region increases steadily, but the tax policyeffects has lots of uncertainties; the government possesses a relatively complete indicatorsystem monitoring tax sources and comparatively abundant statistical data resources on taxeconomy. However, it is difficult to access more precise data for our discussing.Secondly, based on the LS-LM model in the macroeconomics, it studies a mathematicaleconomic model under the principle that economy determines tax and analyses theinfluencing mechanism of factors that determines the tax capacity by using differentialmethod. Combining the literature analysis, it reveals that economic development andeconomic structure factors, for example, the gross domestic products (GDP), consumptionand investment, international trade, market interest rate, money supply, governmentexpenditure demand and so on, determine the size and scale of tax capacity of a country orregion. Besides, political system, policy adjustment and social environment are also the important factors which influence the formation and implementation of tax capacity.According to the available of data, this dissertation builds a model of Chinese regionaltax effort indices selecting the specific influencing factors in aspects of economicdevelopment, economic structure and scale of taxation system scale.Thirdly, the study reveals that there is an inconsistency between the essence ofRegression method’s average estimation and the concept of the maximum potential taxcapacity under the use of Ordinary Least Squares (OLS) method of parameter estimationon regional tax effort index, while there is a similarity between the concept of regional taxeffort indices and production system technology efficiency. The dissertation studies thebasic principle of stochastic simulation technique, proposes the research prospect of theuse of stochastic simulation technique (SFA) to estimate the regional tax effort indices, andputs on sufficient demonstration, comparison and empirical research. It also demonstratesthe validity of measurement result of tax effort indices based on SFA. Additionally, the taxlosing rate of various periods based on the method of SFA possesses higher reliability thanother results of relative researches.Fourthly, the dissertation advances the stochastic hypodissertation of tax base variablesof sub-taxes, and then based on this hypodissertation, it introduces probability distributionmodel and Bayesian’s information revision theory, proposes a relatively uniform idea abouttax capacity measuring distribution estimation algorithm, and then build an estimationintegral model about (absolute) tax capacity of sub-types tax. After that, taking tax capacitymeasuring of salary income tax from individual income tax as an example, it makes anempirical study, which verified the feasibility of the tax capacity measuring distributionestimation algorithm.Finally, during empirical study, based on the data structure characteristics of averageannual salary and numbers of employees in each industry from China Statistical Yearbook,and the prior knowledge about salary differences in internal industrial units, it usesstochastic simulation techniques to generate sample data of employee salary which isapplied to measure individual income tax capacity in China. Thus this research makes agreat contribution to methodology. In a word, this dissertation made lots of work to further study tax capacity measuringmodels and methodologies. The measuring results will be a high contribution in theory andpractice for assisting in adjudging the equity of tax burden, making the transfer paymentdecision and evaluating the local government performance in tax collection.