- Whether the data shows short-term or long-term unemployment, the relationship between unemployment and inflation is tenuous.
- The Phillips Curve has gone through many iterations, including approaches that apply it to expected inflation, different unemployment indicators, and various time frames—but it still doesn’t seem to work that well.
- Alas, just because a relationship exists doesn’t mean it is stable over time or useful for explaining the past, describing the present, or predicting the future.
Initial unemployment claims rose slightly to 320,000 for the week that ended March 15. Continuing claims rose by 41,000 to 2.889 million. Based on labor market patterns over the past 30 months, this is consistent with payrolls growing by 138,000 for March.
The data defies what is known as the Phillips Curve, which correlates low unemployment with high inflation. In 1958, the economist A.W.H. Phillips published a paper that showed a consistent relationship between inflation and unemployment in the U.K. from 1861 to 1957. When inflation was high, unemployment was low, and vice versa. This insight, which continues to be influential today, implies a trade-off for monetary policymakers between inflation and unemployment.
Over certain periods, the Phillips Curve shows some correlation between inflation and unemployment
Source: The Federal Reserve Economic Database
The Phillips Curve: Fact or fable?
Economics is often a form of quantitative storytelling, where data show a trend, theories are constructed to explain it, and stories are told about how policymakers can change variables to affect outcomes. Sometimes, the stories are more like fables. The Phillips Curve seems to be one of those fables.
The Phillips Curve hasn’t been a very good guide for policymakers. It didn’t work during the stagflation of the 1970s, and it doesn’t seem to work today. In all fairness, the Phillips Curve has gone through many iterations, including approaches that apply it to expected inflation, different unemployment indicators, and various time frames—but it still doesn’t seem to work that well.
The limits of the Phillips Curve: more than 65 years of data shows little correlation between inflation and unemployment
Source: The Federal Reserve Economic Database
The New York Federal Reserve published a paper saying that what really matters to inflation is short-term unemployment, not long-term unemployment. Chair Yellen was asked about that research at her first post-policymaking press conference, and she distanced herself from the research, implying that neither she nor most committee members buy this latest iteration of the Phillips Curve. In the research, the authors look at the relationship between short-term unemployment and average compensation costs. They find an inverse relationship.
Alas, just because a relationship exists doesn’t mean it is stable over time or useful for explaining the past, describing the present, or predicting the future. First, average compensation cost and actual price inflation are weakly linked. Inflation, as measured by changes in the Consumer Price Index, depends more on unit labor costs (productivity-adjusted compensation) and credit growth than simple per-hour compensation costs. Unit labor costs haven’t been rising much, and private sector credit creation—especially in the mortgage and credit card departments—hasn’t been growing more rapidly than incomes. As a result, average compensation costs and actual inflation show little connection.
A look at the relationship between short-term versus long-term unemployment and wage pressure
What about the relationship between short-term unemployment and wage pressure? There, again, is a weak link. In fact, changes in long-term unemployment are actually more significant than changes in short-term unemployment. If you look at the relationship between the unemployment rate, broken out by duration (less than 5 weeks, 5 to 14 weeks, 14 to 26 weeks, and 27 weeks or longer), drops in short-term unemployment are indeed weakly correlated to mild increases in future inflation; however, drops in long-term unemployment (27 weeks or longer) are correlated with drops in future inflation. The long-term unemployment figure comprises the biggest fraction of the total unemployed, and that’s the area in which small improvements are likely keeping wage inflation in check.
The Phillips Curve is a workhorse of economics, but life is much more complex than a simple inflation/unemployment trade-off model can capture. Whether you look at short-term unemployment, long-term unemployment, or both, there is little inflationary pressure building.