In 1958 the economist A. W. Phillips dug through nearly a century of British data and spotted a pattern. When unemployment was low, wages and prices rose quickly. When unemployment was high, they barely rose at all. This negative relationship became the Phillips curve — a graph showing that lower unemployment tends to come with higher inflation.
The pattern suggested a tempting menu. The inflation-unemployment trade-off seemed to let policymakers pick their poison. Want lower unemployment? Accept a bit more inflation. Want lower inflation? Tolerate a bit more joblessness. For a while governments treated the curve as a reliable dial they could turn.
The trade-off is real, but only in the short-run Phillips curve. In the short run, a burst of spending lifts demand. Firms hire more and raise prices, so unemployment falls while inflation rises. The catch is that this only works as long as people are fooled about prices.
The hidden ingredient is expectations. Adaptive expectations assume people predict future inflation by looking at the recent past. So a government can surprise them once, but only once. Rational expectations go further, assuming people use all available information, including the government's own playbook, so they are not fooled even briefly. Once workers expect higher inflation, they demand higher wages to match it, and the short-run gains evaporate.
This is why the long-run picture looks completely different. The long-run Phillips curve is a vertical line. Once expectations have fully adjusted, you cannot buy lower unemployment with more inflation. The economy snaps back to a single sustainable level of unemployment no matter how high inflation climbs.
That level has a name. The NAIRU stands for the non-accelerating inflation rate of unemployment, the lowest unemployment the economy can sustain without inflation spiraling upward. Push unemployment below the NAIRU and inflation does not just rise, it keeps accelerating. The NAIRU is essentially the natural rate seen through an inflation lens.
History delivered the proof. The 1970s brought stagflation, with high inflation and high unemployment together, which the original downward-sloping curve said was impossible. The breakdown forced economists to rebuild the model around expectations.
Two lessons followed. First, credibility and forward guidance matter. A central bank that people trust to keep inflation low can steer expectations cheaply. Second came the Lucas critique, the warning that the moment policymakers try to exploit a historical relationship, people change their behavior and the relationship shifts. You cannot ride a curve that moves when you lean on it.