In this paper, we estimate a forward-looking Taylor-type reaction function for China over the period of 2000-2015. In so doing, we use a comprehensive overall policy stance indictor (the Sun-MP index) so as to overcome the measurement uncertainty and hence the model uncertainty problems existing in the current literature, as reflected in various estimations of different rules (both the McCallum and Taylortype rule) for the Chinese case. We find that the Chinese monetary policy is nonlinear. We model the nonlinearity with the multiple-regime threshold regression. With two threshold variables, we identify three regimes featured with economic slowdown, high inflation and Goldilocks economy. This kind of regime-switching model helps tackle regime and model specification uncertainty problems. Our main findings include: The PBC’s policy priority varies across regimes. It fights against low growth and high inflation, but rarely responds to economic overheating or low inflation (e.g., deflation). This bias might be due to the strong belief in China that “growth is panacea”.

This is the abstract of this paper. Please do not cite or distribute without permission of the author.