The New Keynesian Phillips Curve in the United States and the euro area: aggregation bias, stability and robustness

In the recent past, the empirical literature on the New Keynesian Phillips Curve (NKPC) has grown rapidly. The NKPC has been shown to describe satisfactorily the relationship between inflation and marginal cost both for the United States and the euro area. However, little attention has been given so far to the stability and robustness of the parameters in the estimated NKPC. In this paper, we aim to help fill this gap. After estimating hybrid NKPCs on US and euro-area data using the generalised method of moments and having found that our results are broadly in line with previous findings, we subject our estimated NKPCs to a thorough stability analysis. We find that the estimated coefficients for the United States are stable, whereas those for the euro area are considerably less stable. We then investigate the possible reasons for this instability. One explanation, explored using the Andrews' test, is the presence of structural breaks. Another possibility is the presence of an aggregation bias, which we investigate by estimating NKPCs for the three largest euro-area economies: Germany, France and Italy. At this disaggregated level, the fit of the NKPC improves, but the coefficients are still unstable. Furthermore, the disaggregated analysis indicates the presence of structural breaks in the three largest euro-area economies.