David Altig gave a good summary of the “missing inflation” arguments in a post from yesterday. The essence of his argument is, I think, that core inflation is simply not rising because inflationary expectations have not risen (which in turn is due to a very credible Fed promise to keep inflation low), and thus wage pressures have not risen.
In contrast, I tend to think that firmly anchored inflationary expectations are a necessary but not sufficient condition for stable core inflation; a labor market that is strong enough to maintain real wages in the presence of rising consumer prices may well generate rising core inflation even when inflationary expectations are muted, given that non-wage prices are downward-sticky and won’t tend to move down to compensate for higher energy and labor costs.
But David’s further discussion about the validity of focusing attention on the core inflation rate, along with the persuasive evidence provided by Mark Thoma that the core inflation rate does not systematically differ from the overall inflation rate, set me to wondering about another possible piece to the “missing inflation” puzzle (to the degree that such a puzzle still exists).
Take a look at this picture, which is a rough replication of one of Mark’s graphs, and which shows the overall CPI inflation rate (measured as the 2-year average inflation rate) compared to the CPI-core inflation rate.
Mark’s point with this picture was that it’s hard to notice much of a systematic bias in the core rate of inflation; on average, it tends to look like the overall CPI rate of inflation. But another feature of this chart strikes me as interesting: the overall CPI inflation rate seems like a pretty good leading indicator of changes in the core inflation rate.
The following chart explores this connection a bit more clearly. The blue line shows the difference between the overall CPI inflation rate and the core rate; it’s positive when the overall inflation rate is above the core rate, and conversely. Meanwhile, the red line shows the trend in the core CPI inflation rate; a positive reading at any particular point in time indicates that the core inflation rate will rise over the next 6 months compared to the rate 6 months previously, and conversely.
My point is simple: when the overall CPI inflation rate is above the core rate, the core rate tends to rise, though it takes several months. This has happened throughout the Greenspan era of monetary policy, and thus throughout a period of very firmly established inflationary expectations.
I don’t believe that the Fed’s inflation-fighting credibility has changed significantly in recent years; I think it’s as strong as it always has been under Greenspan. So if the current situation adheres to the rest of the Greenspan era, we may well expect the core inflation rate to rise over the coming months.
So in short, let me suggest the following: perhaps the only reason that the core rate hasn’t risen more is because we haven’t waited long enough. If the core rate follows the pattern of the past 15-20 years, Ben Bernanke could have his hands full, with a core rate that will be creeping upward in early 2006, just as the economy may be slowing.