Inflation and unemployment

21 Feb 2011 by Jim Fickett.

Many take it as quite obvious that inflation must remain low, basing their opinion on a few common arguments. But historical data do not support these arguments. We have already seen that inflation expectations do not foreshadow inflation, and that the amount of spare capacity does not correlate with inflation. What about the remaining argument, based on high unemployment? This argument fails as well, when examined on the basis of historical data rather than economic ideals.

Here is a graph of the unemployment rate and the inflation rate, both during the one modern period of high US inflation, 1966-1996:

(Both data series are from the BLS. Click for larger image.)

Since both series are oscillating between low and high values, it is difficult to be sure in which direction any causation may flow. But to me it looks more as if high inflation could lead to high unemployment.

• First, note that the first two peaks in inflation follow minimal drops in unemployment, so it does not seem the low unemployment caused the inflation
• Second, note that quite high unemployment in 1975, dropping back to a still somewhat high level, did not prevent the grand inflation spike of 1980
• Third, note that if the red curve is shifted right a couple years, it matches the blue curve rather well, suggesting that any causation might be from inflation to unemployment rather than the other way around

This is only one period in one country's history, but it does not support the idea that currently high unemployment means inflation must stay low.

Of course I am not the only one who has noticed this. Economist Robert Gordon of Northwestern University wrote in a recent paper,

When plotted in Figure 1 on a scatter diagram from 1960 to 1980 the inflation and unemployment rates are indeed uncorrelated, with a combination of negative and positive correlations that range all over the map. The negative PC [Phillips Curve] tradeoff appeared to be utterly defunct, leading Arthur Okun in 1980 to describe the Phillips Curve as a UFO, or “unidentified flying object.”

Why, then, do so many believe that low capacity utilization or high unemployment must mean low inflation? Because a simple explanation can be given that seems intuitively appealing. “With factories utilizing only 70% of capacity, manufacturers have no pricing power”, or “With unemployment at 9%, workers are afraid to ask for a raise, meaning that wage inflation is very unlikely.”

It all sounds great. Unfortunately, the data do not support these simple stories.