Headline and Core Inflation: A Recent Decoupling

donsutherland1 · 07-17-08, 12:23 PM

In June, 12-month headline inflation soared to 5.0%. Meanwhile, core inflation ticked up ever so gently to 2.4%. The difference amounts to 206.87 basis points, the highest since September 2005 when the spread was 271.10 basis points. The September 2005 spread will likely be surpassed when the July and August Consumer Price Index (CPI) data is released. It is possible that the spread could reach or exceed 300 basis points, especially as the August year-over-year headline inflation rate could be approaching 6%.

This development raises the question as to whether the core inflation figure is as relevant as it once was. In the past, it was argued that food and energy prices are highly volatile and excluding those prices from the overall CPI figure would eliminate the noise caused by such volatility. That argument worked well so long as fluctuations in food and energy prices assured that the core inflation figure would come out close to the headline figure over the long-run.

However, by the early 2000s, headline inflation began to decouple from core inflation. In general, the annual headline inflation figure began to exceed the core figure on a regular basis. During the January 1995 through October 2002 timeframe, in 51/94 (54.3%) of months, the annual headline figure came out less than the annual core figure. As a result, a more aggressive anti-inflation policy could target core inflation. However, from November 2002 through June 2008, the annual headline figure exceeded the core figure in 61/68 (89.7%) of months.

Since April 2004, the divergence between the two inflation figures has accelerated. Early evidence suggests that a pronounced rise in energy prices is likely driving that decoupling.

The divergence between headline and core inflation shows up quite well when headline inflation is indexed for core inflation. In early 2004, one witnesses the beginning of a sustained divergence between the headline and core CPI.

Given the leading role energy price trends appear to have played, it makes sense to examine the price of crude oil. The data reveals that at the time headline inflation was accelerating in its divergence from core inflation, the ascent in the price of crude oil was accelerating.

Right now, it remains too soon to suggest that the price of crude oil will slow its increase for an extended period of time, or even retreat substantially over coming years. The International Energy Agency’s medium-term oil report suggests that a general continuation of a narrow balance between supply and demand. If so, the price of crude oil might well prove far stickier on the upside than it did during the 1980s and 1990s. If that becomes the case, the utility of focusing on the core inflation figure will diminish. Furthermore, a general focus by central banks on the core figure could allow for a persistent widening of the spread between the headline and core rates. In turn, because consumers lack the ability to exclude food and energy from their purchases, tolerance of higher headline inflation from a focus on the core rate could lead to increasing inflation expectations and a growing risk of second order effects.

donsutherland1 · 07-18-08, 08:44 AM

In an interview with reporters from Le Figaro, the Frankfurter Allgemeine Zeitung, the Irish Times, the Jornal de Negócios on July 11, 2008, European Central Bank President Jean-Claude Trichet addressed a question that concerned core inflation. Mr. Trichet explained that the core inflation rate is not "a good predictor of future inflation."

The question and Mr. Trichet's response follow:

Core inflation in the euro area is now about 1.7%, so why should we be so worried about inflation?

First, core inflation is higher than that and secondly, we do not consider core inflation as a good predictor of future inflation. But let me give you an example of why we see signs that justify our concerns. Unit Labour Costs are an important indicator of future inflation, because they are a major cost in the euro area economy. In recent years, we have had an increase rate in Unit Labour Costs of 1.0% in 2005; 0.9% in 2006; 1.5% in 2007; and 2.4% in the first quarter of the present year. When you try to disentangle the factors behind this development, about half of the increase is attributable to increases in wages and salaries and the other half is due to a reduction of labour productivity. The overall result is that we have observed in the last two years a significant rise of unit labour costs.

donsutherland1 · 07-22-08, 08:24 AM

This morning, in an address at the Philadelphia Business Journal Book of Lists Power Breakfast in King of Prussia, Pennsylvania, Philadelphia Federal Reserve Charles Plosser weighed in on the issue of a decoupling of headline and core inflation. He argues that central bankers need to place greater weight on headline inflation even though core inflation remains relatively quiescent. He also notes that trends in headline and PCE (personal consumption expenditures) inflation have better predictive power than changes in the core rate.

Relevant excerpts follow:

In discussing how to make monetary policy decisions, many economists as well as monetary policymakers have tended to focus on the behavior of core inflation rather than headline inflation for several reasons...

Some economists focus on core inflation because a change in the relative price of a key commodity, such as the rise in the price of oil, can affect near-term headline inflation but not necessarily the longer-term trend of inflation. Historically, the prices of food or energy have been quite volatile and at times have temporarily raised or lowered headline measures of inflation, only to settle down to more normal levels over time. Therefore, to have monetary policy react to what may very well be a short-term deviation has usually been considered a mistake...

Moreover, over the last 10 years, headline and core inflation have diverged by 40 to 50 basis points, and thus, the trend in core inflation has not been a reliable indicator of the trend in headline inflation.

Another reason to focus on core inflation would be if core inflation did a better job of forecasting future headline inflation than headline inflation itself. Since policymakers must be forward-looking in making policy decisions, such a forecasting ability would mean that we should monitor core rather than headline inflation.

Although some empirical work on inflation forecasting has supported this view for certain measures of inflation, recent research by our Philadelphia Fed staff suggests that measures of core inflation do not consistently outperform headline inflation in forecasting future headline inflation. That research also finds that more accurate forecasts are obtained by combining the CPI and PCE measures of inflation and that monitoring an array of inflation measures has merit...

Since energy price increases have been so persistent in recent years, I do believe more attention should now be paid to measures of headline inflation in setting monetary policy. I don’t believe we can be sanguine that the behavior of core inflation will keep the public’s inflation expectations well-anchored in the face of persistently high headline inflation.