Macroeconomics › Section L · The Keynesian perspective › Topic L.3

The Phillips curve

For a while in the 1960s, economists thought governments could pick their preferred mix of unemployment and inflation from a simple curve. The Phillips curve explains why that seemed possible, and the 1970s showed why the menu kept changing.

A curved supply curve

The reverse-L supply curve of simple Keynesian analysis is a useful sketch, but research suggests the real short-run aggregate supply curve bends gradually. Far below potential GDP it is almost flat: this is often called the Keynesian zone, where shifts in demand change output but barely move prices. Near and above potential it becomes very steep, the neoclassical zone, where demand mainly moves prices. In between is an intermediate zone, where more demand raises both output and the price level. That intermediate zone is the key to the Phillips curve: as spending pushes the economy towards capacity, unemployment falls but prices start rising faster.

Phillips's discovery

In 1958 A. W. Phillips, a New Zealand-born economist at the London School of Economics, published a study of British data covering about a century, from 1861 to 1957. He found that years of low unemployment tended to be years of rapid growth in money wages, and years of high unemployment brought slow wage growth. Because faster wage growth feeds into faster price growth, economists soon recast the relationship in terms of price inflation. In 1960 Paul Samuelson and Robert Solow sketched a version for the United States using price inflation. The Phillips curve became the name for a downward-sloping link between joblessness and inflation.

Reading a Phillips curve

A Phillips curve puts the unemployment rate on the horizontal axis and the inflation rate on the vertical axis. It slopes downwards: lower unemployment comes with higher inflation, and vice versa. The logic follows from the curved supply curve. When aggregate demand rises, firms produce more and hire more, so unemployment falls. As workers and capacity grow scarce, firms bid up wages and raise prices faster, so inflation rises. When demand falls, the reverse happens. Moving along the Phillips curve tells the same story as moving along the supply curve.

Worked example: reading the menu

A Phillips curve passes through three points: 3% unemployment with 6% inflation, 5% unemployment with 3% inflation, and 7% unemployment with 1.5% inflation. 1. Start at 5% unemployment and 3% inflation. 2. Policy raises demand to cut unemployment to 3%. Reading the curve, inflation rises to 6%. 3. The trade-off is 3 points more inflation for 2 points less unemployment, or 1.5 points of inflation per point of unemployment. 4. Going the other way, from 5% to 7% unemployment, cuts inflation by only 1.5 points, or 0.75 per point. 5. Check: the curve is steeper at low unemployment, matching a supply curve that steepens near capacity.

The 1960s policy menu

During the 1960s many economists and policymakers treated the Phillips curve as a menu. A country could choose low inflation with high unemployment, high inflation with low unemployment, or somewhere in between, and then use fiscal and monetary policy to move along the curve to that point. US data for the 1960s seemed to support the idea: as unemployment fell through the decade, inflation crept up, and the points traced out a fairly neat downward-sloping line. Demand management seemed able to buy lower unemployment at a predictable price.

Stagflation breaks the menu

In the 1970s the neat pattern fell apart. Inflation and unemployment rose together, a combination called stagflation. In 1974 US consumer prices rose by about 11%, and unemployment climbed to 9.0% in May 1975. A second bout of high inflation and high unemployment followed around 1980 to 1982, and many other countries had similar experiences. Points that should have lain on one curve now lay well above and to the right of it. Simple Keynesian analysis, built around shifts in demand, could not explain both rates rising at once. The curve had not been disproved, but it had shifted.

Why the curve shifts

Economists identify two main reasons. The first is supply shocks. When oil prices quadrupled in 1973 to 1974, production costs rose across the economy, shifting the aggregate supply curve to the left: output fell and prices rose, so unemployment and inflation rose together. The second is inflation expectations. If workers and firms come to expect, say, 5% inflation, they build it into wage deals and prices, so any given unemployment rate now comes with more inflation. Both forces shift the short-run Phillips curve. The view that expectations remove any lasting trade-off is developed in the topic on neoclassical policy implications.

Worked example: spotting a shift

An economy has 5% unemployment and 2% inflation. A year later, after a jump in world energy prices, unemployment is 6.5% and inflation is 6%. 1. On a single downward-sloping curve, higher unemployment should come with lower inflation, not higher. 2. Both rates rose, so the economy has not moved along its old curve. 3. The cause is a supply shock: higher energy costs shift aggregate supply left, cutting output and raising prices. 4. Conclusion: the short-run Phillips curve has shifted up and to the right. 5. Check: a fall in demand alone would raise unemployment but lower inflation, so it cannot explain this data.

Keynesian policy along the curve

Keynesian policy aims for the right amount of aggregate demand. In a recession, it recommends expansionary fiscal policy: tax cuts to lift consumption and investment, or more government purchases, shifting AD to the right towards potential GDP. Keynes insisted that in a deep slump the priority is enough spending, even if the projects are not ideal. When the economy runs above potential and inflation is the danger, it recommends contractionary fiscal policy: tax rises or spending cuts to shift AD left, easing price pressure with little loss of output. The goal is a Goldilocks level of demand: not too little, which brings unemployment, and not too much, which brings inflation.

Is the curve still useful?

Since the 1990s the short-run curve has often looked flat. In September 2019, US unemployment fell to 3.5%, its lowest rate since 1969, yet inflation stayed below the Federal Reserve's 2% target. Many economists credit well-anchored expectations. Then in June 2022, after pandemic spending and supply disruptions, the consumer price index was 9.1% higher than a year earlier, the fastest rise since 1981. The lesson most economists draw is that a downward-sloping Phillips curve is a useful guide over a few years, but it shifts with supply shocks and expectations.

Common mistakes to avoid

First, swapping the axes: unemployment goes on the horizontal axis and inflation on the vertical. Second, treating the curve as a fixed menu; it shifts with supply shocks and expectations. Third, explaining stagflation with a fall in demand alone, which would lower inflation, not raise it. Fourth, confusing a movement along the curve, caused by a change in demand, with a shift of the curve, caused by supply shocks or expectations. Fifth, forgetting that Phillips's original curve used money-wage growth, not price inflation.

Key takeaways

A curved aggregate supply curve, flat far below potential and steep near it, implies a downward-sloping Phillips curve: lower unemployment comes with higher inflation in the short run. Phillips found the pattern in British wage data from 1861 to 1957. In the 1960s the curve was treated as a policy menu, but 1970s stagflation showed it shifts, because of supply shocks and inflation expectations. Keynesian policy uses expansionary fiscal policy in recessions and contractionary policy against inflation.

Flashcards

Flashcard

Phillips curve

A downward-sloping relationship between unemployment and inflation

Unemployment on the horizontal axis.

Test your understanding

Question 1 of 5

A Phillips curve passes through 4% unemployment with 5% inflation and 6% unemployment with 2% inflation. If demand falls and unemployment rises from 4% to 6%, what happens to inflation along this curve?

Question 2 of 5

What did A. W. Phillips's original 1958 study relate to unemployment?

Question 3 of 5

Unemployment and inflation both rise sharply after a large jump in world oil prices. What best explains this?

Question 4 of 5

The economy is above potential GDP and inflation is rising. What is the Keynesian policy response?

Question 5 of 5

In the Keynesian zone of a curved aggregate supply curve, what does a rise in aggregate demand mainly change?

← Previous topicNext topic →