Testing the hypothesis of Phillips curve trade-off: a regime switching approach

Abstract

The hypothesis of Phillips curve trade-off is tested by applying a two-state first-order Markov switching model to estimate a simple expectation-augmented Phillips curve, which allows for the possibility that samples are drawn from two normal distributions. Evidence from the UK shows that, two Phillips-curve regimes are identified, and the regime with ineffective trade-off dominates the other over time, which verifies the hypothesis.