The New Keynesian Hybrid Phillips Curve: An Assessment of Competing Specifications for the United States

Inflation forecasting is fundamental to monetary policy. In practice, however, economists are faced with competing goals: accuracy and theoretical consistency. Recent work by Fuhrer and Moore (1995), Galí and Gertler (1999), Galí, Gertler, and Lopez-Salido (2001), Sbordone (2002), and Kozicki and Tinsley (2002a, b) suggests that the two objectives need not be mutually exclusive in the context of inflation forecasts. The New Keynesian Phillips curve is theoretically appealing, because its purely forward-looking specification is based on a model of optimal pricing behaviour with rational expectations. This specification, however, does not properly capture observed inflation persistence. The author estimates three structural models of U.S. inflation that incorporate price frictions to justify the presence of lags in the forward-looking New Keynesian Phillips curve. The models, based on Galí and Gertler (1999) and Kozicki and Tinsley (2002a, b), are tested on the basis of forecast performances. The results show that the new Keynesian hybrid Phillips curve with the output gap as an explanatory variable performs marginally better than the two alternative specifications.