extremely interesting work with such conclusions:
The dynamics of inflation for both the US and Japan may be summarized as follows:
1. Inflation is not well-modeled by an old-style accelerationist Phillips curve, or by a forward-looking, rational expectations New-Keynesian Phillips curve.
2. Inflation may be sensibly modeled as depending on the one-year (survey) expectation of inflation with a weight of between 2/3 and 3/4, and lagged inflation with a weight of between 1/3 and 1/4.
3. Inflation also depends on marginal cost.
4. The one-year expectation responds sluggishly to current and lagged realized inflation rates, and to current and lagged output gaps.
further it is better to follow .
really not worth explaining, what does it mean, if it's closer to the truth, than old school?