Phillips Curve: Unemployment and Inflation Short vs Long Run

Introduction

Phillips observed a consistent inverse relationship between wage inflation and unemployment when he analyzed data from the UK spanning nearly a century, from 1861 to 1957. The explanation Phillips offered was straightforward: the lower the unemployment rate, the more employers had to do to attract talent, and raising wages was one of the primary ways to accomplish that (Wulwick, 1987). In a tight labor market, companies would race to raise wages quickly, while during periods of higher unemployment there would be less pressure to incentivize workers, as the latter would consider themselves fortunate simply to have a job. Since the wages offered to laborers are linked to the prices businesses charge consumers (Lucas & Rapping, 1969), it was not long before economists applied the Phillips curve to inflation in general rather than to wages alone. It became evident that monetary policy could be guided by using the Phillips curve as a model.

One economist who objected, however, was Milton Friedman, who argued that employers and laborers cared only about the purchasing power of real wages, which would adjust so that supply and demand forces could operate freely in the labor market. The rate of unemployment, in other words, was not tied to inflation but rather to real wages. Friedman (1977) objected to the central bank's notion that it could trade higher inflation for lower unemployment. By attempting to manipulate the labor market below the level that natural supply and demand forces would determine, Friedman contended that the government and central bank would create a money illusion — the sense among laborers that they were being paid more — when in reality their money had less purchasing power because of inflation.

Friedman distinguished between the short-run and the long-run of the Phillips curve to make his point. This paper will show why Friedman was right to raise his objection.

Short-Run vs. Long-Run Phillips Curve

The difference between the short-run and the long-run with respect to the Phillips curve, as demonstrated by Friedman (1977), is that when the natural rate of inflation is generally constant — as it was during the 1960s — the inverse relationship between inflation and unemployment holds true. This is the short-run. However, when the central bank attempts to adjust the rate of inflation in order to lower unemployment, the long-run shows that unemployment will return to its natural rate once the labor market processes the price inflation caused by central bank intervention. The attempt to alter a natural process (the unemployment rate) through an artificial adjustment of a correlated variable does not produce a long-term effect on that natural process, but it does produce a long-term effect on the inflation rate. Friedman (1977) demonstrated in the 1970s that the sooner workers adjusted their expectations of price inflation, the less successful the government would be in applying monetary policy to the problem of unemployment.

Friedman's argument led to the creation of the expectations-augmented Phillips curve and the consensus view that at some point a stable rate of unemployment can be found that is consistent with a stable rate of inflation. However, the data show that the relationship tends to be more complicated in reality than in theory (Stiglitz, 1997). An assessment and analysis of 20 years of U.S. unemployment and inflation data will illustrate this complexity. The relationship between unemployment and inflation differs in the short-run — wherein inflation rises and unemployment drops — from the long-run — wherein inflation persists but unemployment returns to its natural rate — because of the money illusion factor in the short-run and the eventual recognition by laborers of the diminished value of their real wages.

Twenty-Year U.S. Unemployment and Inflation Data

An assessment of 20 years of U.S. unemployment and inflation data (1999 to 2019) does confirm the short-run Phillips curve in certain periods. In 1999, unemployment stood at 4.0% and inflation was at 2.7%. In 2000, inflation rose to 3.4% and unemployment dropped slightly to 3.9%. In 2001, inflation fell to 1.6% and unemployment rose to 5.7% (Amadeo, 2019). Over the short-run, therefore, one can observe a correlation between inflation rate movement and unemployment rate movement. However, by 2002, the pattern broke down. Inflation rose to 2.4% and unemployment also rose to 6.0%. Inflation then fell back to 1.9% and unemployment likewise fell to 5.7% in 2003. Inflation rose the following two years to 3.4% while unemployment trended downward to 4.9%. Inflation fell in 2006, only to rise again in 2007, with unemployment mirroring the movement rather than inverting it.

In 2008, the Federal Reserve slashed rates aggressively and inflation fell to 0.1% from 4.1% the prior year, while unemployment rose from 5.0% to 7.3%. Inflation rose to 2.7% in 2009 and unemployment also rose to 9.9% that same year. Only twice from that point forward — in 2011 and in 2016 — was there an inverse relationship between inflation and unemployment (Amadeo, 2019).

The long-term FRED chart from the St. Louis Federal Reserve illustrates the long-run picture more clearly. From 2003 onward, the 10-Year Breakeven Inflation Rate hovered around 2.5%, dropping sharply to near 0% in 2008 before rebounding the following year, reaching 2.5% again, and then trending downward within a range of 1.6% to 2.11% from 2014 onward. Unemployment, however, went from 4.6% when inflation was near zero to 9.9% when it rebounded to its prior norm of approximately 2.5%. As the inflation rate has modestly trended downward since then, the unemployment rate has fallen more dramatically, reaching approximately 3.0% (FRED, 2019).

Conclusion

By intervening in the markets through the manipulation of inflation rates in order to reduce unemployment, the central bank produces an effect opposite of what it intends: it accelerates inflation while doing effectively nothing to alter the long-run course of unemployment. Only short-run alterations may occur, but over time the labor market recognizes that the purchasing power of wages has declined, and the unemployment rate naturally adjusts as supply and demand restore equilibrium. Today, the Phillips curve has little relevance as a tool for central banks, as there has been no statistically significant correlation between inflation rate movement and unemployment rate movement over the past 20 years — and especially not since 2008, when the central bank's use of unconventional monetary policy drove an upswing across asset classes and correlated with rising service prices across industries. What instrument can the central bank use to moderate unemployment? Its hope rests in further QE — but whether that hope is well or ill founded remains to be seen.

References

Amadeo, K. (2019). Unemployment rate by year since 1929 compared to inflation and GDP. Retrieved from https://www.thebalance.com/unemployment-rate-by-year-3305506

FRED. (2019). FRED Graph. Retrieved from https://fred.stlouisfed.org/series/LNU04000024#0

Friedman, M. (1977). Nobel lecture: inflation and unemployment. Journal of Political Economy, 85(3), 451–472.

Heller, R. (2017). Monetary mischief and the debt trap. Cato Journal, 37(2), 247–261.

Lucas, R. E., & Rapping, L. A. (1969). Price expectations and the Phillips curve. The American Economic Review, 59(3), 342–350.

Stiglitz, J. (1997). Reflections on the natural rate hypothesis. Journal of Economic Perspectives, 11(1), 3–10.

Wulwick, N. J. (1987). The Phillips curve: which? whose? to do what? how? Southern Economic Journal, 834–857.