What is the Phillips curve?

The Phillips curve, named after economist A.W. Phillips, describes an empirical inverse relationship between the unemployment rate and the rate of wage (and by extension, price) inflation. The original observation from 1958 showed that when unemployment is low, workers have bargaining power and wages rise, pushing up costs and eventually consumer prices. When unemployment is high, the reverse occurs.

The relationship has been modified substantially since its original formulation. Milton Friedman and Edmund Phelps introduced the "expectations-augmented" Phillips curve in the late 1960s, arguing that the tradeoff between unemployment and inflation is only temporary. In the long run, inflation expectations adjust, and the economy gravitates toward a "natural rate of unemployment" (NAIRU) regardless of the inflation rate. This insight explained the stagflation of the 1970s, when both unemployment and inflation rose simultaneously.

The modern Phillips curve used in Federal Reserve models relates the output gap (or unemployment gap) to inflation after controlling for inflation expectations and supply shocks. This version has appeared "flat" for much of the past 25 years, meaning that changes in unemployment had only modest effects on inflation. During the 2010s, unemployment fell from 10% to 3.5% with minimal inflation, leading many economists to question whether the Phillips curve was "dead."

The 2021-2023 period revived the Phillips curve debate. The extremely tight labor market (unemployment below 3.5%, job openings at record levels) coincided with the highest inflation in 40 years, exactly as a steep Phillips curve would predict. However, disentangling demand-driven inflation from supply-chain disruptions and fiscal stimulus made it difficult to isolate the pure labor market channel. The Fed's framework still incorporates a Phillips curve relationship, but with deep uncertainty about its slope, position, and stability over time.