Modeling Inflation After the Crisis

Working Paper: NBER ID: w16488

Authors: James H. Stock; Mark W. Watson

Abstract: In the United States, the rate of price inflation falls in recessions. Turning this observation into a useful inflation forecasting equation is difficult because of multiple sources of time variation in the inflation process, including changes in Fed policy and credibility. We propose a tightly parameterized model in which the deviation of inflation from a stochastic trend (which we interpret as long-term expected inflation) reacts stably to a new gap measure, which we call the unemployment recession gap. The short-term response of inflation to an increase in this gap is stable, but the long-term response depends on the resilience, or anchoring, of trend inflation. Dynamic simulations (given the path of unemployment) match the paths of inflation during post-1960 downturns, including the current one.

Keywords: Inflation; Unemployment; Recession; Forecasting

JEL Codes: C22; E31


Causal Claims Network Graph

Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.


Causal Claims

CauseEffect
unemployment recession gap (J64)inflation (E31)
unemployment recession gap (J64)inflation (E31)

Back to index