Why can't the Fed just fix unemployment and inflation at the same time?
Background
Phillips published his scatter plot in 1958. Within a decade it had become the central operating framework for stabilization policy in most advanced economies. The stagflation of the 1970s destroyed the simple version and installed the expectations-augmented model in its place.
The modern version is more conditional: the tradeoff exists in the short run when expectations are anchored, the curve's slope depends on credibility and the size of shocks, and the natural rate of unemployment is not fixed -- it drifts with labor market structure, technology, and demographics.
What it covers
A.W. Phillips documented a negative relationship between wage inflation and unemployment in U.K. data in 1958. Samuelson and Solow quickly translated it into a policy menu: policymakers could, in principle, choose a point on the curve -- accepting more inflation to get less unemployment.
Friedman and Phelps independently dismantled that menu in 1968. If workers and firms adjust their expectations to whatever inflation the central bank produces, the tradeoff disappears in the long run. The economy returns to the natural rate of unemployment -- the NAIRU -- at higher inflation but no better employment. The long-run Phillips curve is vertical.
The 2021-2022 inflation episode reopened every empirical question about the curve's slope, the stability of expectations anchoring, and whether the pre-pandemic NAIRU estimates had any meaning.
Open question
Why can't a central bank permanently buy lower unemployment by accepting higher inflation?