Phillips Curve

What Is the Phillips Curve?

The Phillips Curve is the most influential, and most controversial, relationship in macroeconomics: the empirical observation that unemployment and inflation tend to move in opposite directions. When the labour market is tight (low unemployment), inflation rises. When the labour market is slack (high unemployment), inflation falls.

Named after New Zealand economist A.W. Phillips, who documented the relationship between UK unemployment and wage growth in a 1958 paper, the Phillips Curve became the intellectual foundation of modern central banking. Every time the Fed decides whether to raise or cut interest rates, it is implicitly making a judgment about the Phillips Curve, whether the current level of unemployment is generating too much, too little, or just the right amount of inflation pressure.

For traders, the Phillips Curve matters because it determines how the Fed interprets labour market data. A strong NFP report or a declining unemployment rate isn't just economic news, it's an inflation signal that feeds directly into the Fed's reaction function and therefore into every interest rate, equity, and currency trade.

The Original Phillips Curve

Phillips's 1958 Finding

Phillips analysed 97 years of British data (1861-1957) and found a stable, inverse relationship between unemployment and wage inflation. The curve was non-linear:

• At very low unemployment (<3%), wage growth accelerated sharply, a steep portion of the curve
• At moderate unemployment (4-6%), wage growth was modest, a flat portion
• At high unemployment (>7%), wages stagnated or fell

Samuelson and Solow's Extension (1960)

American economists Paul Samuelson and Robert Solow adapted Phillips's wage curve into a price inflation curve, arguing that wage increases translate into price increases as firms pass through higher labour costs. They presented policymakers with a seemingly straightforward menu:

This "menu" implied that policymakers could choose their preferred combination, and for a decade, this framework guided economic policy.

The Evolution: From Simple to Complex

Stage 1: The Original Trade-Off (1960s)

The 1960s seemed to validate the Phillips Curve. Kennedy and Johnson administration economists used it to justify expansionary policies: push unemployment down, accept a bit more inflation, and enjoy the prosperity. Unemployment fell from 7% (1961) to 3.4% (1969). Inflation rose from 1% to 6%. The curve worked, or seemed to.

Stage 2: Stagflation Destroys the Simple Curve (1970s)

The 1970s shattered the original Phillips Curve. Both unemployment AND inflation rose simultaneously, a combination the curve said was impossible:

What went wrong: The original Phillips Curve ignored inflation expectations. When workers and businesses expect higher inflation, they demand higher wages and raise prices pre-emptively, shifting the entire curve upward. The 1970s oil shocks combined with un-anchored inflation expectations produced a new, higher Phillips Curve at every unemployment level.

Stage 3: The Expectations-Augmented Phillips Curve (1970s-1990s)

Milton Friedman and Edmund Phelps independently predicted the 1970s breakdown. Their key insight: the Phillips Curve trade-off is only temporary. In the long run, unemployment returns to its "natural rate" (NAIRU) regardless of inflation. Attempts to permanently lower unemployment below NAIRU simply produce ever-accelerating inflation.

The expectations-augmented Phillips Curve:

Inflation = Expected Inflation − β(Unemployment − NAIRU) + Supply Shock

This framework says:

• Inflation equals expected inflation when unemployment is at NAIRU (no trade-off in the long run)
• Inflation rises above expectations when unemployment is below NAIRU
• Inflation falls below expectations when unemployment is above NAIRU
• Supply shocks (oil prices, supply chains) can shift inflation independent of the labour market

This became the dominant framework at central banks and remains the intellectual basis of the Fed's approach today.

Stage 4: The Flat Phillips Curve (2000-2019)

From the mid-1990s through 2019, the Phillips Curve appeared to flatten to near zero, unemployment fell from 6% to 3.5% over two decades, yet core inflation barely budged from 1.5-2.0%. The "β" coefficient (sensitivity of inflation to unemployment) shrank toward zero in every empirical estimate.

Why it flattened:

The policy consequence: The flat Phillips Curve gave the Fed license to keep rates near zero for years. If low unemployment doesn't cause inflation, there's no reason to raise rates pre-emptively. This contributed to the extended zero-rate period of 2008-2015 and the cautious hiking cycle of 2015-2018.

Stage 5: The Non-Linear Revival (2020-Present)

COVID revealed that the Phillips Curve is not flat, it's non-linear:

• Flat at moderate unemployment (3.5-6%): The curve is nearly horizontal. Inflation is insensitive to unemployment changes in this range. This is what prevailed in 2010-2019.
• Steep at extreme tightness (below ~3.5%): When unemployment drops well below NAIRU and the labour market is extremely tight, inflation surges. This is what happened in 2021-2022 when unemployment fell to 3.4%.
• Steep at extreme weakness (above ~8%): During severe recessions, deflation risks emerge as mass unemployment destroys demand.

The non-linear Phillips Curve reconciles the 2010s (flat, because unemployment was in the moderate range) with 2021-2022 (steep, because unemployment was at historic lows).

NAIRU: The Equilibrium Unemployment Rate

What Is NAIRU?

NAIRU (Non-Accelerating Inflation Rate of Unemployment) is the unemployment rate at which inflation is stable. Below NAIRU, inflation accelerates. Above NAIRU, inflation decelerates. At NAIRU, inflation stays constant (at whatever level expectations are anchored).

NAIRU is the Phillips Curve's equilibrium point, and like the neutral interest rate (r*), it is unobservable, time-varying, and surrounded by enormous uncertainty.

NAIRU Estimates Over Time

The NAIRU Uncertainty Problem

The Fed's NAIRU estimate has a confidence interval of approximately ±1.5 percentage points. If NAIRU is 4.0%, unemployment of 3.5% is only slightly inflationary. If NAIRU is 5.5%, unemployment of 3.5% is a dangerously overheated labour market. The difference between these two scenarios is the difference between holding rates steady and hiking aggressively.

This uncertainty is why the Fed has moved toward "data-dependent" policy rather than targeting a specific unemployment rate. They watch inflation itself, not just the labour market, because NAIRU is too imprecise to rely upon exclusively.

Trading the Phillips Curve

The Labour Market → Inflation → Rates Chain

Every labour market data release triggers a Phillips Curve calculation in the market's mind:

The Average Hourly Earnings Trade

The most direct Phillips Curve trade focuses on the Average Hourly Earnings (AHE) component of the NFP report:

• AHE MoM > 0.4%: Wage inflation accelerating → Phillips Curve pressure → hawkish → sell 2Y Treasuries, buy USD
• AHE MoM 0.2-0.3%: Consistent with target → no Phillips Curve alarm → neutral
• AHE MoM < 0.2%: Wage growth decelerating → Phillips Curve easing → dovish → buy 2Y Treasuries, sell USD

When the Phillips Curve Breaks Down (Trading the Surprise)

The highest-value trades occur when the Phillips Curve doesn't work as expected:

1. Low unemployment + falling inflation (2023-2024): Supply-side improvement → Phillips Curve shifts left → the Fed can be more dovish than expected → buy risk assets aggressively
2. Rising unemployment + persistent inflation (stagflation): Phillips Curve fails entirely → the Fed has no good option → buy gold, sell both stocks and bonds
3. Low unemployment + stable low inflation (2017-2019): Phillips Curve is flat → the Fed can stay easy → buy growth stocks and duration

What to Watch

1. NFP Average Hourly Earnings: The most direct Phillips Curve variable. Track MoM and YoY wage growth relative to expectations.
2. JOLTS job openings-to-unemployed ratio: A broader measure of labour market tightness than unemployment alone. A ratio above 1.5 signals extreme tightness.
3. Atlanta Fed Wage Growth Tracker: A more stable measure of individual wage growth (tracks the same individuals over time), reducing composition effects.
4. FOMC "longer-run unemployment rate": The Fed's NAIRU estimate, published in the SEP. Shifts signal changing Phillips Curve assumptions.
5. Unit labour costs: Wages adjusted for productivity. Rising unit labour costs = inflation pressure even if productivity is growing. This is the "true" Phillips Curve variable that central banks care about.