Inflation Regimes and the Sources of Inflation Uncertainty
利用马尔可夫转换模型构建美国战后通胀不确定性指标，发现其能解释调查预期，且体制不确定性成分对预测失业有显著作用。
This paper develops new time series measures of inflation uncertainty in the United States in the postwar period that account for the prospect of changing inflation regimes. The measures are constructed from estimates of a Markov switching model for inflation. Importantly, we show that rational forecasts derived from the Markov model are consistent with survey measures of inflation expectations. Our Markov model allows us to decompose uncertainty about future inflation into two components; a certainty equivalent component that ignores uncertainty about future inflation regimes, and a regime uncertainty component that reflects this uncertainty. Survey measures of inflation uncertainty, based on the dispersion of forecasts, appear more closely associated to the regime uncertainty component than the certainty equivalent component of inflation uncertainty. The regime uncertainty component also appears to have significant explanatory power in forecasting unemployment while the certainty equivalent component does not. Copyright 1993 by Ohio State University Press.