The Bones and Shapes of the Phillips Curve

TL;DR

This paper investigates whether the Phillips curve's relationship between inflation and unemployment has fundamentally changed or simply nonlinear during and after the COVID-19 pandemic, using advanced econometric methods on US data.

Contribution

It introduces a threshold model to distinguish between structural slope stability and non-linearities in the Phillips curve during the pandemic period.

Findings

No significant change in the Phillips curve slope post-pandemic

Identification of non-linear activation in inflation dynamics during labor market tightness

Insights into unemployment costs of disinflation during extreme economic conditions

Abstract

The COVID-19 pandemic reignited debate on the U.S. Phillips curve. Using MSA-level panel data (2001-2024), we employ a Two-Stage Least Squares (2SLS) instrumental variable strategy with a shift-share instrument to estimate core non-tradable inflation's response to a v/u-based slack measure. We distinguish structural slope stability from state-dependent non-linearities via a threshold model. Our analysis addresses whether the slope of the Phillips Curve changed during and after the Pandemic in the United States by evaluating if recent inflation dynamics reflect an altered structural trade-off ("bones") or the activation of non-linear "shapes" in response to extreme labor market tightness. This distinction offers critical insights into the unemployment cost of disinflation.