The new Keynesian Phillips curve: endogeneity and misspecification

Abstract

The wrong or insignificant sign of the forcing variable in the new Keynesian Phillips curve estimations may be a result of the endogeneity of the labour share and misspecification of real marginal cost in the baseline model. We address the misspecification of real marginal cost by formulating a broad measure that features the labour share, output gap and supply shock variables. The endogeneity of the labour share is addressed by using an appropriate lag of the labour share in the Phillips curve. Reduced-form evidence from five developed and five emerging market economies support the empirical validity of the NKPC.