Abstract: To measure reasonably inflation uncertainty and identify properly its sources have great policy implication for the government to manage effectively inflation expectation. This paper applies time-varying parameter model with Markov regime switching Heteroskedasticity to decompose inflation uncertainty into structural uncertainty and impulse uncertainty, and to investigate which one is dominant, then to test their impact on the volatility of macroeconomic. The result shows that impulse uncertainty played the dominant role and the structural one decreased after 1996, which can be well explained by the traits of the economic transformation, the evolvement of macro-control and the external shocks. And the further analysis finds that different sources of inflation uncertainty have different impacts on macroeconomic activities. Further analysis find impulse uncertainty increases the volatility of output growth. However structural uncertainty slows down the volatility of output growth, but this effect is not obvious.

Key words: Inflation Uncertainty, Expectation Management, Markov Regime Switching Model