Federal Reserve’s Chairman Jerome Powell testified last month before the House Committee on Financial Services. As usual, Powell answered questions regarding the present and future of monetary policy.
During the testimony, Democratic Congresswoman Alexandria Ocasio-Cortez asked Powell about the weakening of Phillips curve and its relevance for today’s monetary-policy decisions. Ocasio-Cortez’s comments on the Phillips curve were unexpectedly praised by President Trump’s top economic advisor Larry Kudlow, who pointed out that the Phillips curve was dead. But what is the Phillips curve?
In the late 1950s, William Phillips, a New Zealand economist, identified an inverse relationship between unemployment and nominal wage growth using UK data: lower unemployment rates were correlated with higher wage inflation, and vice versa. This negative correlation was immediately interpreted as a trade-off: when governments try to reduce unemployment via fiscal or monetary policy, unemployment falls and nominal wages increase, pushing up labor costs and ultimately the price level.
The figure below shows this inverse relationship, aka the Phillips curve. Governments can lower the unemployment rate, but at the expense of higher inflation (the economy moves from point A to point B). Yet, as Milton Friedman and Edmund Phelps showed in the late 1960s, this trade-off is only temporary. In the long run, the Phillips curve is vertical: there is no relationship at all between unemployment and inflation.
The apparent trade-off results from economic agents lacking expectations about inflation. Once expectations are formed, this relationship vanishes. Imagine that government increases public spending dramatically. The new money entering the economy pushes prices up, incentivizing companies to hire new workers as revenues increase, thereby reducing the unemployment rate. We would move from point A to point B of the above graph.
Yet this situation doesn’t last long. At the new inflation rate, the purchasing power of workers is lower than before. In addition, workers now anticipate higher inflation, which lead them to demand higher wages to make up for the loss in purchasing power. This reestablishes the previous unemployment level, but at a higher inflation rate. We would now be in point C. If the process keeps repeating over and over again, we get an inflationary spiral: increasing inflation with no long-term effects on the unemployment rate.
If we assume rational expectations, however, this trade-off shouldn’t exist even in the short term as economic agents factor in all relevant information when generating inflation expectations. Under rational expectations, economic agents are able to anticipate the effects of monetary and fiscal policy in advance, which prevents workers from being fooled into accepting lower real wages in the first place.
Nonetheless, the inverse relationship between inflation and unemployment continues to hold in the short term, probably due to the existence of rigidities in the economy. Particularly, labor contracts prevent wages from adjusting automatically when inflation expectations rise, creating the trade-off described above.
What is true is that the unemployment-inflation relationship has gradually weakened in the US over the last three decades, hence Kudlow’s and Ocasio-Cortez’ comments about the death of the Phillips curve. This weakening has been attributed to factors such as globalization, decrease in the bargaining power of workers, and central banks’ credible commitment to price stability.
In any case, it should be noted that, contrary to conventional wisdom, we cannot infer from the Phillips curve that lower unemployment resulting from economic growth necessarily creates inflation. Economic growth implies an increased production of goods and services, which, all else equal, increases real wages and puts downward pressure on prices.
Instead, the corollary is that monetary and fiscal policy can only get the unemployment rate back to its natural rate (the level of employment compatible with a stable inflation rate), which doesn’t depend on monetary/fiscal stimuli but on institutional factors such as flexibility of the labor market or the duration of unemployment benefits, among others.
Or simply put, in the long run, governments and central banks can’t draw upon money printing and public spending to achieve lower unemployment rates. The only thing they can do is to bolster political reforms aimed at unleashing the power of markets to generate economic growth through innovation and competition.
 In practice, high inflation rates tend to go hand in hand with productivity-undermining government policies, which move the long-term Phillips curve to the right. The result is high inflation combined with increasing unemployment.