The New Keynesian Phillips Curve for a small open economy

Abstract

Abstract:\nThe New Keynesian Phillips Curve (NKPC) has become the benchmark model for understanding\ninflation in modern monetary economics. One reason for the popularity is the microfoundation of the\nmodel, which decomposes agents' behaviour into price adjustments and deviations of the price level\nfrom its target. The empirical relevance of the NKPC is, however, a matter of debate as recent\nstudies reveal that some supportive evidence depends crucially on the econometric methods applied.\nWe show how to evaluate the features of the model using cointegration techniques and tests based\non both single-behavioural equations and cointegrated VAR models. Our results indicate that the\nforward-looking part of the NKPC is most likely at odds with Norwegian data. By contrast, we\nestablish a well-specified dynamic model interpreted as a standard backward-looking mark-up price\nequation. We also demonstrate that the dynamic mark-up model forecasts well post-sample and\nduring a major change in the monetary policy regime, which certainly is strong evidence in favour of\nthis model. Consequently, we conclude that taking account of forward-looking behaviour when\nmodelling consumer price inflation in Norway seems unnecessary to arrive at a well-specified model\nby econometric criteria.\nKeywords: The New Keynesian Phillips Curve, mark-up pricing, single-equation estimation\nmethods, encompassing tests, cointegrated vector autoregressive models and equilibrium correction\nmodels.