Research On China’s Optimal Flexible And Time-varying Fiscal Policy Response Function Based On Flexible Loss Function

Fiscal policy is an important foundation and pillar of national governance,and fiscal authorities need to use fiscal policy to smooth economic dynamics and promote economic growth.As a mathematical expression of fiscal policy rules,the fiscal policy response function is not only the focus of research by policy makers and economic scholars,but also the focus of economic participants.Currently,domestic and foreign scholars mainly set constant coefficients or time-varying fiscal policy response functions,but some scholars have suggested that there is an asymmetric phenomenon in China’s fiscal policy adjustment process,and most of the fiscal policy response functions are currently set symmetrically.Moreover,in recent years,various uncertainties have been frequent and the macroeconomic environment has become more uncertain,so some scholars have suggested that the fiscal policy response function should be flexible.It is obvious that there is an urgent need to improve the flexibility of the traditional constant coefficient or time-varying fiscal policy response function,and to explore the form and asymmetric characteristics of the optimal flexible time-varying fiscal policy response function.Therefore,the study of the optimal flexible time-varying fiscal policy response function has certain theoretical and practical significance,enriches the theory of loss functions of fiscal authorities and the theory of the mechanism of fiscal policy rules,and provides suggestions and guidance for fiscal authorities to study and formulate fiscal policies.This paper first constructs a flexible loss function with multiple forms on the basis of domestic and foreign research results,and obtains a flexible time-varying fiscal policy response function by deriving the flexible loss function and adding time-varying coefficients.The model with the lowest mean squared error and standard deviation in the estimation results of the mixed-frequency mixed innovation time-varying coefficient random variance regression model is chosen as the optimal form of fiscal policy response function,and the statistical and trend characteristics of its coefficient estimates are analyzed to draw conclusions and give relevant recommendations on macroeconomic operations.Chapter 1 provides the background and significance of this paper and a literature review;Chapter 2 introduces the fiscal policy reaction function through the debate on camera choice and fiscal rule,briefly introduces the types and evolution of fiscal policy reaction function,and gives a theoretical framework for the study of flexible and timevarying fiscal policy reaction function;Chapter 3,by sorting out the forms of fiscal policy reaction function at home and abroad,and focusing on the current fiscal policy In Chapter 4,a mixed-frequency mixed innovation time-varying coefficient random variance regression model is used to empirically estimate the optimal time-varying fiscal policy reaction function and to analyses the relevant empirical results.The study concludes that,firstly,the constant coefficient model tends to overestimate the deficit rate smoothing coefficient,the time-varying model tends to underestimate the deficit rate smoothing coefficient and has greater volatility,and the flexible time-varying deficit rate smoothing coefficient is optimal in terms of size,stability and flexibility;secondly,the response of the fiscal deficit rate to changes in the debt gap in the optimal model is characterized by symmetric inert interval preference with a wide span of flexible time-varying and exhibit pro-cyclical characteristics;third,the response of the fiscal deficit to changes in the output gap exhibits flexible time-varying asymmetric preference characteristics,and China’s fiscal policy is more sensitive to the positive output gap than the negative output gap.Innovations of this paper: A loss function with flexibility in terms of inertia interval and symmetry is constructed,and the optimal flexible time-varying fiscal policy response function is obtained by deriving the flexible loss function;a mixedfrequency mixed innovation time-varying coefficient random variance regression model is used for mixed-frequency regression analysis.Weaknesses: compared to the DSGE approach,it lacks a micro-foundation;and only three policy variables,namely deficit rate,debt rate gap and output gap,are selected and the interrelationship between the three variables is neglected.