Phillips Curve Flattening

The Phillips curve has flattened over the past three decades: the tight inverse relationship between unemployment and inflation that defined postwar macroeconomics broke down. Tighter labor markets no longer reliably push inflation higher, reshaping monetary policy strategy and expectations of when the central bank will act.

The original Phillips curve and its breakdown

In the 1950s–1960s, economist A.W. Phillips observed a stable negative relationship: lower unemployment consistently coincided with higher inflation. Policymakers embraced the idea: you could choose a point on the curve, trading off joblessness against rising prices. Tight labor markets meant wage pressure; wage pressure meant cost-push inflation.

The relationship held through the 1960s. But in the 1970s it collapsed. Stagflation—high inflation and high unemployment simultaneously—appeared impossible under the old curve but became the decade’s defining crisis. The curve had shifted outward.

Economists eventually understood why: inflation expectations had become unanchored. When workers and firms expected prices to rise, they demanded higher wages and set higher prices preemptively, decoupling inflation from current unemployment. The Phillips curve was not stable; it shifted with expectations.

Why the curve flattened after the 1990s

The 1980s brought a second lesson. Federal Reserve chair Paul Volcker’s aggressive rate hikes killed the wage-price spiral by convincing markets that the central bank would not tolerate inflation. Expectations reanchored. By the 1990s, survey inflation expectations had stabilized around 2–2.5 percent.

Once expectations were firmly anchored, the Phillips curve did not vanish—but it flattened. A 1 percentage point drop in unemployment no longer reliably raised inflation by 0.5 percentage points as it had in the 1960s. Instead, it might raise inflation by 0.1 or 0.2 percentage points, or barely at all.

The mechanism: with expectations stable, firms and workers do not automatically demand higher prices when unemployment falls. Wage growth rises more slowly. Price markups remain contained. Inflation responds weakly to slack, not strongly.

Mechanisms behind the flattening

Multiple overlapping forces have kept inflation subdued despite tight labor markets:

Expectations stability forms the foundation. Forward guidance from the Federal Reserve—explicit promises to keep rates low or raise them gradually—anchored long-term inflation expectations near 2 percent, weakening the link from tight labor to wage pressure.

Globalization reduced the ability of domestic labor markets to push wages and prices higher. When firms can source labor or production overseas, tight domestic unemployment does not immediately raise wage pressure. Global supply chains and offshoring act as a price anchor.

Technology and productivity growth, particularly in low-wage services, held margins and prices down even as unemployment fell. Firms could meet demand without raising wages proportionally.

Weak bargaining power among workers, despite low unemployment, kept wage growth moderate. Union membership declined. Gig work and temporary arrangements reduced workers’ leverage in negotiation.

Supply-side increases in labor supply—immigration, longer work-life participation—reduced effective tightness even at headline unemployment rates reported by unemployment statistics.

Measurement debates: What counts as slack?

A subtler debate concerns whether the Phillips curve truly flattened or whether headline unemployment is a poor measure of labor market tightness.

Unemployment alone is a narrow metric. It misses underemployment—workers in part-time jobs seeking full-time work. It excludes discouraged workers who left the labor force. As recessions and long-term joblessness rose, effective slack may have been much larger than the unemployment rate suggested, keeping inflation benign despite the headline number falling.

Some economists argue the Phillips curve relationship remains intact once you measure slack broadly—including wage growth, underemployment, and labor force participation—rather than relying only on the unemployment rate. Others counter that the relationship has genuinely become looser and less predictable.

Implications for monetary policy

The flattened Phillips curve changed monetary policy calculus. In the 1960s and early 1970s, the curve was steep; policymakers faced a hard trade-off: buy a big unemployment reduction and get big inflation. Now the curve is flat; you can reduce unemployment significantly with modest inflation consequences, at least in normal times.

This raised a possibility: perhaps the long-run natural rate of unemployment was lower than previously thought, and the Federal Reserve could sustain lower joblessness without igniting inflation. This conviction drove policy choices in the late 2010s, as the Fed maintained low rates while unemployment fell below 4 percent without seeing the inflation surge once feared.

But the flattened curve also created a blind spot. When inflation does emerge—as it did in 2021–2022 with supply shocks and demand stimulus—a flat curve can create overshoot. Small labor market tightness may trigger large inflation if expectations begin to shift or supply is severely constrained. The relationship is not absent; it is nonlinear and conditional.

Recent evidence and ongoing debates

The post-2008 period tested these ideas. Despite the unemployment rate falling from 10 percent to near 3.5 percent by 2019, inflation remained subdued. The flat Phillips curve prediction held. But 2021–2023 added wrinkles. Inflation spiked despite unemployment still being low, driven by supply chain disruptions, fiscal stimulus, and global factors. Wage growth did accelerate in tight labor markets, but it lagged inflation, reducing real purchasing power.

This suggests the Phillips curve did not disappear entirely, but the relationship is weaker, slower, and more conditional on expectations and supply constraints than the vintage 1960s curve. Economists continue to debate whether the flattening is permanent or cyclical—a product of the low-inflation regime that could shift if inflation expectations truly unanchor.

See also