Working Paper

The Phillips Curve's and Relative Phillips Curve's Slopes: Why So Different?

Abstract: I estimate the effect of labor market tightness on wage inflation from 2004-2019 using aggregate data and a hybrid New Keynesian Phillips curve. The Phillips curve slope, i.e., the effect of a unit increase in the vacancy-unemployment ratio on inflation, is about 3.4 percentage points. Then, I estimate the model using the corresponding panel-level data with a time-fixed-effect regression: The resulting regional (i.e., relative) Phillips curve slope is about 0.7. This large difference between the two slopes is robust to controlling for various measures of inflation expectations and for supply shocks. The gap arises because variation used in one regression is—by construction—orthogonal that used in the other. I explain how cross-region spillovers might explain the large gap between the aggregate and relative slopes.

JEL Classification: E31;

https://doi.org/10.20955/wp.2025.010

Bibliographic Information

Provider: Federal Reserve Bank of St. Louis

Part of Series: Working Papers

Publication Date: 2025-05-18

Number: 2025-010