|
| |
| Kuo-Hsuan Chin and Xin-Hua Zheng |
| |
| ''Stability of Phillips Curve: The case of Taiwan'' |
| ( 2024, Vol. 44 No.2 ) |
| |
| |
| We study the time-varying slope of the Phillips curve by applying a tri-variate hybrid VAR model with time-varying parameter and stochastic volatility (hybrid TVP-VAR-SV, hereafter) to Taiwan's macroeconomic data. We follow Chan and Eisenstat (2018b) by using a Bayesian approach to approximate the posterior density of (time-varying) parameters and the marginal likelihood of a model, in which the latter one is used for the model comparison. We find that the fitness of a hybrid TVP-VAR-SV model to Taiwan's data is superior to a “non-hybrid” TVP-VAR-SV model of Primiceri (2005), a widely-used one in the current studies. In particular, the estimated parameters shown in the inflation equation, found in the best model and used to characterize the Phillips curve, are time-invariant, supporting the stability of Taiwan's Phillips curve over the past four decades. Moreover, we also find the stability of Phillips curve for China, the largest trading partner for Taiwan and similar to Taiwan in both language and culture, over the past two decades. |
| |
| |
| Keywords: Bayesian Approach, Phillips Curve, Time Varying Parameter, Vector Autoregression. |
JEL: E3 - Prices, Business Fluctuations, and Cycles: General (includes Measurement and Data)
C1 - Econometric and Statistical Methods: General |
| |
| Manuscript Received : Jul 20 2023 | | Manuscript Accepted : Jun 30 2024 |
|
