The Phillips Curve

In 1958, a New Zealand-born engineer turned economist named A.W. Phillips published a paper that would reshape how governments thought about managing their economies. Working at the London School of Economics, Phillips plotted nearly a century of British data --- wage inflation against unemployment from 1861 to 1957 --- and found a remarkably stable inverse relationship. When unemployment was low, wages rose quickly. When unemployment was high, wages barely moved. The curve that fit this data became one of the most famous diagrams in economics.

From Wages to Prices: The American Translation

Phillips himself was cautious about his findings, but two towering American economists were not. In 1960, Paul Samuelson and Robert Solow took the Phillips Curve to the United States, replacing wage inflation with price inflation on the vertical axis. Their version offered policymakers something intoxicating: a menu of choices. Want lower unemployment? Accept a bit more inflation. Prefer stable prices? Tolerate higher joblessness. The economy, it seemed, presented a permanent tradeoff that governments could navigate by turning the dials of fiscal and monetary policy.

This was a powerful idea for the Keynesian consensus of the 1960s. The Kennedy and Johnson administrations embraced it enthusiastically, using tax cuts and spending programs to push unemployment down while accepting modest inflation as the price. For a time, it appeared to work beautifully. The 1960s saw unemployment fall steadily while inflation remained contained --- the curve seemed to be a reliable guide.

The Friedman-Phelps Revolution

The comfortable policy menu did not survive the decade. In his 1967 presidential address to the American Economic Association, Milton Friedman launched a devastating theoretical attack. Edmund Phelps, working independently, reached similar conclusions. Their argument was elegant and, in retrospect, almost obvious.

Workers and firms are not stupid, Friedman argued. If the government tries to exploit the Phillips Curve by running the economy hot, people will eventually notice that prices are rising and adjust their expectations accordingly. Workers will demand higher wages to compensate for expected inflation. Firms will raise prices to cover those higher wages. The result: inflation accelerates, but unemployment returns to its original level --- what Friedman called the “natural rate.”

This meant the long-run Phillips Curve was not a downward-sloping menu at all. It was vertical. In the long run, there was no tradeoff between inflation and unemployment. The only way to keep unemployment below its natural rate was to continuously accelerate inflation, surprising workers again and again --- an obviously unsustainable strategy. The “expectations-augmented” Phillips Curve replaced the original, adding expected inflation as a shifting factor that moved the short-run curve upward over time.

Stagflation: The Empirical Blow

If Friedman and Phelps delivered the theoretical critique, the 1970s delivered the empirical one. The oil shocks of 1973 and 1979, combined with years of expansionary policy, produced something the original Phillips Curve said should not exist: stagflation. Inflation and unemployment rose simultaneously. The neat inverse relationship collapsed in the data, and with it, the confidence of a generation of policymakers.

The stagflation episode was a watershed. It discredited the simple Keynesian version of the Phillips Curve, elevated monetarism and rational expectations theory, and contributed to the political shift toward the anti-inflation policies of Paul Volcker’s Federal Reserve and the governments of Margaret Thatcher and Ronald Reagan. The lesson drawn was stark: trying to buy permanently lower unemployment with a little inflation was a fool’s errand.

New Keynesian Reformulation

Economics rarely discards an idea entirely --- it refurbishes it. By the 1980s and 1990s, New Keynesian economists had rebuilt the Phillips Curve on more rigorous foundations. In their models, firms face “menu costs” and other frictions that prevent them from adjusting prices instantly. Wages are set in staggered contracts. These real-world frictions mean that monetary policy can affect output and unemployment in the short run, even if people have rational expectations.

The New Keynesian Phillips Curve relates current inflation to expected future inflation and a measure of economic slack (often the output gap). It preserves the long-run neutrality that Friedman insisted on --- money cannot permanently push unemployment below its natural rate --- while explaining why monetary policy has real short-run effects. This framework became the backbone of inflation targeting by central banks worldwide.

The Flattening Puzzle

After the 2008 financial crisis, the Phillips Curve threw another curveball. Unemployment in the United States soared to 10 percent, yet inflation barely fell. When unemployment later dropped to historic lows before the pandemic, inflation barely rose. The relationship, if it still existed, appeared to have flattened dramatically. Some economists declared the Phillips Curve dead.

Several explanations emerged. Globalization may have weakened the link between domestic labor markets and prices. Well-anchored inflation expectations, a product of credible central banking, might have reduced the sensitivity of inflation to unemployment. The rise of the gig economy, declining union power, and increased market concentration all potentially altered how labor market tightness translates into wage and price pressure.

Then came the pandemic. The massive fiscal and monetary response of 2020-2021, combined with supply chain disruptions, produced the sharpest inflation surge in four decades. Suddenly, the Phillips Curve seemed alive again --- or at least, the broader principle that overheating an economy eventually produces inflation reasserted itself.

The Enduring Lesson

The Phillips Curve’s history is a case study in how economics progresses. An empirical regularity gets elevated to a policy tool. The tool breaks because it ignores how people adapt their behavior. A more sophisticated version emerges, incorporating those adaptations. And the debate continues, because the underlying question --- what is the relationship between real economic activity and inflation? --- is one that every central banker and finance minister must answer, whether they want to or not.

The honest answer, after seven decades, is that the relationship is real but unstable, context-dependent, and far more complex than a single curve on a graph. That is less satisfying than the 1960s policy menu, but it is closer to the truth.