The Phillips Curve

In 1958, economist William Phillips plotted decades of British unemployment against wage growth and found a clean, negative relationship: when unemployment was low, wages rose fast; when unemployment was high, wage growth stalled. The intuition is straightforward — a tight labor market forces employers to compete for workers by paying more, and that cost eventually shows up in the prices they charge. That relationship, generalized from wages to inflation broadly, is the Phillips curve.

The basic shape and why it moves

Plotted as a curve, low unemployment sits on the left with high inflation, high unemployment sits on the right with low or falling inflation. A simple version:

$$\pi = \pi^e - \beta(u - u^*)$$

In words: actual inflation $\pi$ equals expected inflation $\pi^e$, adjusted down by how far unemployment $u$ sits above the "natural" rate $u^*$ the economy can sustain without accelerating inflation, scaled by how sensitive prices are to labor market slack, $\beta$. When unemployment is below that natural rate, the second term goes negative, and inflation runs hotter than expected; when it's above, inflation runs cooler.

Worked example

Suppose the natural rate of unemployment $u^*$ is 4.5%, $\beta = 0.6$, and expected inflation is 2%. If actual unemployment falls to 3.5% — one percentage point below the natural rate — predicted inflation is $\pi = 2\% - 0.6 \times (3.5\% - 4.5\%) = 2\% + 0.6\% = 2.6\%$. If unemployment instead rises to 6.0%, 1.5 points above the natural rate, predicted inflation falls to $\pi = 2\% - 0.6 \times (6.0\% - 4.5\%) = 2\% - 0.9\% = 1.1\%$. The same model, same expected inflation, produces very different outcomes purely from where unemployment sits relative to that natural-rate benchmark.

What this means in practice

Central banks lean on the Phillips curve logic when they raise rates to cool an overheating labor market and bring inflation down, accepting some rise in unemployment as the cost. But the relationship has weakened and flattened in many developed economies since the 1990s — inflation has stayed relatively stable through periods of both very low and moderately high unemployment, which is one reason policymakers now talk about a "flat" Phillips curve and rely more heavily on inflation expectations as a separate input.