Phillips Curve

The Phillips Curve describes an inverse relationship: when unemployment is low, inflation tends to rise, and vice versa. First documented in 1958 by economist A.W. Phillips, this relationship became the intellectual foundation of post-war monetary policy—and one of the most contested concepts in modern macroeconomics after stagflation shattered its predictive power in the 1970s.

The original Phillips observation

In 1958, economist A.W. Phillips published a deceptively simple finding: he had charted 97 years of British wage growth against the unemployment rate and found a clear downward-sloping curve. When joblessness was high, wage growth languished; when tight labour markets pushed unemployment toward 2%, wages surged. The relationship held across different eras and even different countries.

The intuition appealed to policymakers immediately. A central bank facing a choice between unemployment and inflation could exploit the tradeoff: tolerate slightly more joblessness to keep prices stable, or accept inflation to bring unemployment down. The Phillips Curve implied this choice existed in the short run.

Most early applications focused on wage growth rather than price inflation, but the logic extended: higher wages lead to higher production costs, which firms pass through to customers. Wage-price spirals were the mechanism. Tight labour markets bid up wages; higher wages push up prices; the cycle accelerates unless central banks tighten policy.

The expectations-augmented Phillips Curve

The original model’s pristine downward slope broke apart in the 1970s. Stagflation—simultaneous high unemployment and high inflation—plagued the U.S., UK, and other advanced economies. The Phillips Curve seemed to vanish, or at least shifted outward in ways the original framework did not predict.

Economists Milton Friedman and Edmund Phelps independently proposed a crucial refinement: the Phillips Curve should account for inflation expectations. Workers and firms do not bargain over real (inflation-adjusted) wages blindly; they anticipate future inflation and embed that into wage demands. A central bank cannot permanently trade unemployment for inflation without continuously surprising the public with higher inflation than expected—a trick that works only once.

The expectations-augmented Phillips Curve reads:

Inflation = Expected inflation + f(Unemployment gap) + Supply shocks

When unemployment falls below its “natural” or non-accelerating-inflation rate (NAIRU), inflation accelerates. When unemployment rises above NAIRU, inflation decelerates. But the intercept—the rate of inflation when unemployment is at NAIRU—depends on what people expect inflation to be going forward.

This version explained the 1970s. Inflation expectations had become untethered from monetary policy as the Federal Reserve pursued expansionary policy. Workers came to expect high inflation and bargained for it, pushing the entire Phillips Curve upward. The central bank faced a crueller choice: accept stagflation or break expectations through a painful recession.

The flatness puzzle of recent decades

A new puzzle emerged after the 1990s. The Phillips Curve appeared to flatten: unemployment fell sharply from 2000 to 2007, yet inflation remained subdued. Conventional estimates would have predicted a much stronger inflation response. Again, the relationship seemed broken.

Several explanations compete. Globalisation (cheap imports from Asia) may have disciplined wage growth and price-setting. Central banks gained credibility and anchored inflation expectations, so firms and workers no longer expected high inflation even when joblessness fell. Quantitative easing and low interest rates kept financial conditions loose without directly tightening labour markets as much as they historically would have.

An alternative view holds that the Phillips Curve never flattened—it just shifted. As long-term inflation expectations remained well-anchored (thanks to credible central banks), the relationship between current unemployment and current inflation weakened, because people were not as surprised by low inflation. The Phillips Curve still exists, but operates conditional on stable expectations.

The 2021–2023 test

The post-pandemic inflation episode reignited the debate. From 2020 to 2023, unemployment fell sharply and inflation surged across advanced economies. Some economists pointed to this as a vindication of the Phillips Curve; others noted that supply shocks (semiconductor shortages, energy commodities spikes, shipping disruptions) explained much of the inflation, not labour market tightness alone.

The controversy centres on a simple question: How much does tight labour demand actually push up wages and inflation? If the effect is weak (as recent flatness suggested), then expansionary policy is safer and can focus on unemployment. If the effect is strong (as 2021–2023 seemed to show), central banks must act sooner to head off inflation.

Most recent evidence points to a nonlinear Phillips Curve: the relationship is weak when unemployment is high (slack is abundant) but strengthens sharply as unemployment approaches very low levels. This would reconcile the apparent flatness of the 2000s-2010s with the inflation acceleration of 2021–2023.

Policy implications today

Central banks now estimate unemployment gaps and inflation expectations with sophisticated models, then use a Phillips Curve framework (however modified) to forecast inflation and justify policy moves. The Federal Reserve explicitly targets maximum employment alongside price stability, making the Phillips Curve tradeoff central to its forward guidance.

Disagreement persists. Some economists (often associated with “dovish” policy) argue the Phillips Curve is so flat that unemployment can be brought low without igniting inflation. Others (hawk-leaning) see the relationship as more dangerous and argue central banks must raise rates preemptively to keep inflation expectations anchored. The Phillips Curve framework does not resolve this debate; it simply structures it.

See also