| The Phillips Curve that bridges inflation and output gap has important academic implications and research value.A detailed study of its shape and the factors that influence it will be of great significance to the resilience of China’s economic development and the scientific formulation of macroeconomic policies.This paper examines the variation in the shape of the Phillips curve from two perspectives: the first is a series of studies on the shape of the curve itself,and the second is a study of the impact of policy regulation on the shape of the curve.However,an important prerequisite for this series of studies is the accurate measurement of the key influencing variable output gap,and as there is no academically recognised output gap in China,this paper attempts to construct an econometric evaluation system to select the most appropriate output gap for China’s economic reality.The main research work carried out in this paper in turn is as follows.Firstly,this paper illustrates the mathematical basis and construction principles of the commonly used econometric models under four categories of output gap measurement methods.It measures the output gap under 14 methods including the production function method,QT filter,DT model,DT-20 model,UCUR-20 model,UCUR-(09,20)model,BN decomposition,HP filter,BK filter,CF filter,wavelet decomposition,SVAR model and bivariate dynamic system filter model as well as the pseudo-output gap,and their morphological characteristics were initially characterised.It is found that,except for the wavelet gap and the pseudo-output gap,which have significantly different trends,the trends of the results of the remaining methods remain basically consistent,which indicates that the gaps can be regarded as robustness tests for each other and confirm the trends of the output gap,but the vast majority of the gaps suffer from the drawback of being difficult to accurately reflect key economic facts,and only the UCUR-(09,20)gap and the pseudo-output gap can match the typicalised facts of our economic performance to a greater extent.Secondly,this paper designs an econometric evaluation system for output gaps in five dimensions: smoothness,ability to fit economic facts,coupling association,forecasting ability and measurement robustness,to achieve the selection of output gaps for the actual macroeconomic operation of China.It is found that the UCUR-(09,20)gap is outstanding in terms of smoothness,ability to fit economic facts and coupling,and is relatively superior in terms of forecasting ability and measurement robustness,and the fuzzy comprehensive evaluation results show that it has the highest score and is the optimal measurement choice that best reflects the actual operation of the output gap in China.The pseudo-output gap has deep economic connotations,but underperforms in the dimension of forecasting ability,which may lead to the estimation bias of the Phillips curve.The common filtering methods represented by the HP filter have obvious shortcomings in some dimensions,and comprehensive capability evaluation show that they are not a satisfactory measure for estimating the output gap.Thirdly,this paper further eliminates the possible estimation bias caused by the theoretical model setting and the choice of inflation proxy variables and achieves the estimation of the Phillips curve shape in China.The GMM estimation results show that the most suitable combination to portray the shape of the Phillips curve in China is "CPI inflation-UCUR-(09,20)gap-supply shock augmented New Keynesian hybrid Phillips curve model",the Phillips curve for China still exists and the slope is estimated to be 0.1119.The GMM estimation results for the combination suggest that using the wrong output gap will lead to a misjudgement that the Phillips curve association has disappeared or has flattened.By controlling for inflation and output gap states,research into the identification of conditions for the flattening of the Phillips curve shows that pure economic slack does not lead to a flattening of the curve,but when low inflation and economic slack coincide,the curve will take on a flattening character,and when the economy is experiencing low inflation,there is a need to be highly alert to signals of economic slack such as insufficient effective demand and weakening expectations.Next,this paper applies the previous Phillips curve measurement framework to identify the heterogeneous features of the regional Phillips curves and compares them with those of the national Phillips curve.Results show that the national curve based on the new sample is still present and significant,confirming the robustness of the previous measurement framework,but the curve estimates for each region are significantly different from the national curve,showing a divergence of flattening in the east,inverting in the mid-west and steepening in the northeast.A clear geographical dependence of inflation expectations also exists.In order to further identify the objective existence of regional differences,this paper applies the quantile time-frequency connectedness method to estimate the inter-regional connectedness of inflation,output gap and Phillips curve.The inter-regional connectedness of inflation,output gap as well as Phillips curve are largely asymmetric in the time domain,frequency domain and quantile dimensions.In particular,it is noted that when the economy is in recession,the presence of connectedness leads to a further flattening of the Phillips curve,exposing the economy to higher downside risks,while when the economy is expanding,the presence of connectivity inhibits the flattening of the Phillips curve,thus contributing to a further improvement in economic fundamentals.Finally,this paper considers the role of interest rate policy in regulating for inflation,the output gap and the Phillips curve shape respectively.Based on the application of GMM to the estimation of the optimal monetary policy response function,the paper further derives the shape of the loss function of interest rate policy regulation for each component and plots the moderate welfare loss region for each component.Results show that the model parameter estimates based on the UCUR-(09,20)output gap are more significant,again demonstrating the consensuality of this optimal output gap.The welfare loss calculation shows that when the positive(negative)inflation gap does not exceed 0.57%(0.83%)and the output gap does not exceed 0.25%(0.57%),the welfare cost is minimal and the central bank hardly regulates.Once the deviation exceeds these thresholds,the cost of losses rises significantly and the central bank will take appropriate policy action.This paper further applies the QARDL model to explore the effectiveness of interest rate policy in curve shape correction,finding that moderate or higher interest rate regimes are more conducive to improving the slope of the Phillips curve,but that moderate interest rate regime typically has a higher speed of adjustment to equilibrium and more significant coefficient estimates,suggesting that a moderate interest rate level of[2.16%,5.11%] is the most favourable policy horizon for safeguarding the Phillips curve linkage. |
PDF Full Text Request