September 24, 2020

An assessment of the Phillips curve over time: evidence for the United States and the euro area

We assess the stability of the coefficient on the unemployment gap in various linear dynamic Phillips curve models. We allow the coefficient on the unemployment gap and the other variables in our model to be time-varying, so that we can monitor the importance of the Phillips curve over time. We compare the effects of different measures for inflation and inflation expectations on our estimation results. In our analysis, we use state space methods and adopt a practical approach to Bayesian estimation with feasible testing and diagnostic checking procedures. Empirical results are presented for the United States and the five largest euro area economies.

Our main conclusion is that in the United States the Phillips curve for headline inflation has remained empirically relevant over the years while there are periods when its impact has been low. For measures of core inflation we find a declining Phillips curve. In the euro area the strength of the relationship differs per country and over time, but has overall been weak and volatile in the past three decades. For both the United States and the euro area countries, we find little evidence of the “anchored expectations"-hypothesis.

External author: Siem Jan Koopmans (VU)

Doi: https://doi.org/10.34932/9sxf-1876

Contacts

Foto Marente Vlekke
Marente Vlekke +31 6 25687697 Read more
Foto Martin Mellens
Martin Mellens +31 6 31795524 Read more

Read more about