Empirical Research On Sustainbility Of Local Government Debt

In 2008,the state implemented a four trillion investment rescue plan,and local governments need to bear 2.28 trillion of them.Under the restriction that the Budget Law does not allow local governments to borrow,local financing platforms are almost the only option for local governments.With the expansion of the scale of debt raised by financing platforms and the continued easing of the financial credit environment,the stock risk caused by debt accumulation has gradually increased.In 2013,the National Audit Office conducted a comprehensive audit of government debt in 31 provinces of China in accordance with the principle of "seeing people,seeing accounts,and seeing things,one by one,item by item",and had a clear grasp of the local government debt stock.Although local government debt plays an important role in regional economic development,the ever-expanding debt scale will also place a heavy burden on local economic development So,does local government debt affect the operation and sustainability of finances?Correspondingly,what are the measures of local debt sustainability?How to evaluate and predict the sustainability of local debt is worthy of systematic research.This article is the research on this theme.This article is divided into six parts.The first part is an introduction,which introduces the research background and meaning,the interpretation of related concepts,the research content and structure,research methods,and innovations and deficiencies.The second part is a review of the literature,combing the literature from the perspective of the concept,judgment criteria,and empirical strategies of government debt sustainability.The third part analyzes the basic facts of local government debt in China,and briefly introduces the scale,sustainability level and influencing factors of local government debt The fourth part is an empirical analysis of the sustainability assessment of China's local government debt.Starting from the fiscal response function,using the data regression of each province to obtain the equation to solve the theoretical borrowing ceiling of each province.This serves as a measure of the sustainability of local government debt in each province.In addition,a stress test was conducted to analyze the changes in the debt sustainability of local governments in different provinces under different scenarios.The empirical results show that the debt sustainability level of most provinces is still high,but some western provinces such as Qinghai and Gansu have debt debt rates that have exceeded the theoretical borrowing ceiling.Under the premise that the central government does not implement the rescue principle or faces debt unsustainability Debt replacement is conducive to reducing the cost of debt repayment by local governments,which in turn increases the theoretical ceiling of local government borrowing.As the downward pressure on the economy increases,the theoretical ceiling of local government borrowing becomes smaller and smaller,and as the economic growth rate further shrinks The increasing contraction of the theoretical debt ceiling may bring about debt sustainability issues.The fifth part is the prediction of local government debt sustainability based on the random forest model.This part constructs a sustainable prediction input index system that includes the three links of debt borrowing and repayment,and uses the random forest algorithm to train the debt sustainability prediction input.The non-linear relationship between indicators and output indicators,with the help of the trained model,further predicts and analyzes the local government debt sustainability in the next three years.The results show that the sustainability of most provinces is maintained at a medium level,but the specific values have declined.The debt sustainability of the three northeast provinces and western provinces such as Gansu,Qinghai,and Ningxia is not high,and it needs to be paid attention to.Chapter 6 is conclusions and policy recommendations.Based on the research conclusions of this article,the following policy recommendations are proposed:(1)strengthen supervision,clean up and transform local investment and financing platforms;(2)improve the division of central and local expenditure responsibilities,and improve local finance Ability;(3)sort out the correct concept of performance and reduce irrational debt;(4)improve the repayment mechanism and establish a sustainability monitoring system.The innovations of this paper are:(1)In terms of method,this paper chooses to use the random forest method in the study of local government debt sustainability less,and uses machine learning to learn the non-linear leading relationship of debt sustainability input and output data,making The prediction of future government debt sustainability is more accurate.(2)In the construction of the sustainability forecasting index system,the existing research "debt as a debt" has been avoided,and the three links of debt borrowing,use and repayment have been fully considered,and the supervision of The role of improving debt sustainability makes the setting of indicators more scientific.This paper has the following deficiencies:(1)In terms of the local government debt stock data,because the official did not publish the provincial local government debt stock data over the years,the data estimated in this article may be somewhat different from the real debt stock;(2)In the selection of continuous forecast indicators,due to the trade-off between the comprehensiveness and simplicity of the indicators,it is impossible to cover all the indicators related to debt.