The Phillips Curve is dead—long live the Phillips Curve

Questioning the Phillips Curve
Questioning the Phillips Curve
Image: Reuters/Erin Scott
We may earn a commission from links on this page.

A small miracle happened in Washington this week. Larry Kudlow, president Donald Trump’s director of the National Economic Council, and Representative Alexandra Ocasio-Cortez, the firebrand freshman Democrat, were in agreement.

Their common enemy is the Phillips Curve, an old economic theory that guides monetary policy and suggests there must be a trade-off between low inflation and low unemployment. Ocasio-Cortez grilled Federal Reserve chairman Jerome Powell about that relationship, arguing the curve should be abandoned and policies historically associated with causing inflation, like raising the minimum wage (which is not a Fed policy), should be reconsidered. While Powell didn’t touch the minimum wage question, he agreed that lately there is little correlation between low employment and high inflation. Kudlow went further and said that if the Phillips Curve no longer applies, there is room for another interest rate cut.

The Phillips Curve was born in 1958, when New Zealand economist W.H. Phillips noticed that whenever inflation was up, unemployment was down, or at least it was in the UK between 1861 to 1957. The curve effectively died in the 1970s when both inflation and unemployment soared, leading to a major rethink on the relationship between inflation and unemployment, led by economists Milton Friedman and Ned Phelps (I was Phelps’ student in graduate school). The result was a more nuanced understanding of inflation and unemployment. The economy has changed since the 1970s, and while the Phillips Curve may be due for yet another incarnation, that doesn’t make the case for lowering interest rates or other inflationary policies.

Phillips Curve 2.0 presumes there’s a trade-off between the difference between total and structural unemployment, and the difference between actual and expected inflation. Structural unemployment, also called the natural rate, is a poorly understood concept by many people. Ocasio-Cortez called it long-term unemployment, but that’s not the right way to think about it because the structural rate changes over time. It reflects the structural capacity of the economy, the skill levels of workers, where they live, and the wages they are willing to work for. All these things can and do change. But it is important to note these factors aren’t influenced by the Fed’s monetary policy. Cyclical unemployment—or unemployment that depends on the business cycle—is sensitive to monetary policy, which can boost the demand for workers when employers respond to lower interest rates by expanding and hiring. The distinction is important because Fed policy can’t control structural unemployment. If it cuts rates too low and unemployment falls below the structural rate, the Phillips Curve suggests inflation will run faster than people expect as economic activity heats up and more money chases the same number of economic goods and workers. This is normally when the Fed will pull back and raise rates. But for some reason this isn’t happening. Despite the unemployment rate at record lows, inflation has been stable.

One problem is that it is impossible to know precisely how much unemployment is structural and how much is cyclical, as Powell admitted.  The fact that unemployment is 3.7% and there’s still low inflation could mean the structural rate changed. It’s also possible inflation is exceeding expectations, but the difference in yields between nominal and real 5-year bonds—one measure of expectations used by economists—is just 1.8%,  slightly lower than the actual rate of inflation, which is about 2%.

But the question remains if structural unemployment has changed, or if we should junk the Philips Curve. It is possible the Philips Curve relationships have changed. The Fed now has an explicit inflation target in order to create stable expectations, which it didn’t do in the 1970s, so it’s not surprising the relationships between expectations, inflation, and unemployment are different now than prior to inflation targeting. But this doesn’t necessarily mean the Fed should cut rates.  A changing relationship suggests a need for caution; and with historically low unemployment and vigorous GDP growth, it is hard to argue the economy needs more lift.  Low rates pose both costs and benefits. It also worth noting that the Fed mandate is starting to include financial stability, which low rates may undermine.

Fed doves like Kudlow argue both that the Philips Curve is broken and the Fed can cut rates until there’s inflation—but both of those things can’t be true. You can’t argue the Philips Curve justifies a rate cut and also say that it’s dead.