|
--Stone & McCarthy (Princeton)--
The FOMC continues to wrestle with Core Measures of inflation that are above the so-called "Comfort-Zone", while the prospects for economic activity appear to be downshifting to a trajectory, which appears to be comfortably below long-term potential.
The Committee has exhibited understandable patience in tolerating what they believe to be an elevated inflation rate, in part, because members anticipate a softening of Core inflationary measures over the quarters ahead.
But, many on the Committee worry that tolerating an elevated inflation rate for too long, risks embedding higher inflationary expectations into household and business decisions. If such were to occur, elevated inflationary expectations may require more a painful future policy prescription, than preventive medicine now.
It is for this reason that the stated primary concern of policy-makers remains on the inflation front.
Is this a Reasonable Approach to Policy?
Certainly the restraint that the Committee has exercised in not being overly sympathetic to the downside risks to the economy is understandable. Especially, given the heightened awareness of the role of inflationary expectations.
But, one has to ask whether the yardsticks by which the Committee gauges acceptable inflation are appropriate or not. Furthermore, one has to reflect upon whether or not measured inflation such as the Core CPI or Core PCE deflators appropriately gauge underlying inflationary trends, or whether these measures are vulnerable to distortions that may actually be in the opposite direction of underlying inflationary or deflationary forces.
In this article we argue that some flexibility in the evaluation of core-inflation and the comfort-zone is appropriate, setting the stage for what we think will be a less restrictive policy in the not-too-distance future.
Is a 1% to 2% Comfort-Zone an Appropriate Range for Core Inflation?
Maybe, Maybe Not, Maybe the zone should be viewed as time-varying much like other concepts such as NAIRU or Potential GDP?
Our understanding of the history of the Comfort-Zone stems from a discussion that occurred at a 1994 FOMC meeting. At that time, then Fed Governor Janet Yellen asked Fed Chairman Greenspan to quantify the concept of price stability. Greenspan apparently responded by saying price stability is achieved when the CPI, properly measured, is flat over time. Yellen apparently responded by saying that means 1% to 2% when the CPI is improperly measured. From this exchange a "Comfort-Zone" was born.
In the years that followed this concept was refined. During 2003, a period in which the Fed was quite concerned about deflation the 1% to 2% zone was justified, by regarding the first 1% as an insurance policy against deflation, and another 1% as reflecting bias inherent in consumer price measures.
The point is that this Comfort-Zone has become almost an implicit inflation target for the FOMC, and a consistent violation of this zone risks de-anchoring inflationary expectations. That said, this zone is not a well-calculated, precise, delineation of appropriate bounds of inflation, but more a seat-of-the-pants or rule-of-thumb set of inflation boundaries.
John Berry made a significant observation regarding the 1% to 2% "Comfort-Zone". That is that during 2003 when core inflation fell towards 1%, the lower bound of the zone, Chairman Greenspan and then Governor Bernanke weren't especially comfortable, fearing an episode of deflation (See: Fed Far from inflation-Target Policy: John Berry, Nov 21, 2006). Berry raised the issue whether a 1-1/2% to 2-1/2% Core PCE deflator zone might actually be more comfortable.
For sure, if the current behavior of the Core PCE deflator was materially above the 1% to 2% Comfort-Zone, there would be no question that policy restraint would be the appropriate stance.
But, against the backdrop of quite uncertain economic prospects, does the Q4-06 Core PCE deflator standing only 1/4% above the upper bound of the zone, warrant drawing unquestionable policy inferences? Particularly when the zone is somewhat arbitrary, and our measures of Core inflation are somewhat imprecise.
A Key Problem With Core CPI and Core PCE Deflators
The Core CPI and PCE Deflators suffer distortion associated with the index for the cost of owner occupied housing, the so-called Owners Equivalent Rent (OER) component. The extent of the problem in the CPI is somewhat more pronounced simply because the weight of the OER component in the CPI is greater than in the PCE deflator, but the underlying story is the same.
We have addressed these problems in past commentaries. For example, we outlined the broad issues at had in CPI--The Owners Equivalent Rent Debate (June 15, 2006). We discussed how falling utilities costs had the perverse affect of boosting OER and Core consumer prices in Understanding the CPI--Hurricane Katrina and OER (June 21, 2006).
Here we update the OER story since mid-2006. The OER component of the CPI has accounted for about 1/2 of the total gain in Core CPI over the past year, even through the weight of this component is about only 23% per the headline CPI, and about 30% of the Core CPI.
In fact, the acceleration in the Core CPI over the past year has been very much dominated by the OER component. In the 12 months ending February 2007 the Core CPI rose 2.7%, up 0.6% from the 2.1% 12-month increase ending February 2006. Of the 0.6% acceleration in Core CPI, 0.5% is directly attributable to the OER component.
During this same period the OER component accelerated from a 2.5% 12-month increase ending February 2006 to a 4.2% increase in February 2007.
The OER component of the CPI is an attempt to measure the cost of owner occupied housing, not by accounting for cash flows associated with mortgage rates, real estate taxes, or home prices, but rather by accounting for the so-called "opportunity cost" in living in one's own home. That "opportunity cost" is defined as the rent one would receive if they rented their home to someone else.
The OER is by design an imperfect measure of owner occupied housing costs, but in the late 1970s and early 1980s when owner occupied housing costs were measured by the cash flows there were a host of other problems with this CPI component.
The primary input into the OER accounting is the behavior of rents on housing units that are similar to owner occupied units. For cities such as NYC, apartment rents serve as a proxy for owner occupied condo type units. The BLS makes some adjustment to rents in estimating OER to account for units in which utilities are included in the basic rental charge. This is more common in places like NYC, than in some other regions of the country.
The debate over the past year has focused upon the relationship between the behavior of home prices and rents. When the OER concept was developed the thinking was that rents and home prices would be closely aligned. But, in recent years the behavior of home prices and rents have been inversely related.
From early 2002 to early 2005 home prices accelerated sharply, while rents generally moderated. It was during this period that the OER contribution to Core CPI diminished, falling from about 1.4% to 0.7%, accounting for much of the deceleration in Core CPI, and adding to the deflationary fears that existed for a time.
More recently home prices have slowed in a material fashion. In the 12 months ending December the Case-Shiller composite index for 10 major cities slowed to only +0.5% from +16.0% in the 12 months ending December 2005. While home prices were slowing dramatically rents began rising sharply. In the 12-months ending February residential rents had risen 4.6%, up from 3.1% in the period ending February 2006.
Why have home prices and rents switched from exhibiting similar patterns to exhibiting an inverse relationship in recent years? There are a number of reasons. But, the most common explanation is simply that in a period of falling home prices prospective occupants would rather rent than buy, inducing an increase in the demand for rental units, and an associated rise in rents.
Clearly falling home prices should be regarded as a deflationary phenomenon. This was certainly the situation during the Great Depression, and in Japan in the 1990s. That said, the rising contribution from the OER component of CPI is an artifact of moderating home prices.
Perversely a restrictive monetary policy, which may induce further moderation in home prices, could have the net affect of boosting OER and ultimately Core CPI.
What is the Outlook for the OER?
All indications suggest that the OER will likely continue to make a disproportionate contribution to Core CPI and the Core PCE deflator. There appears little reason to anticipate a sudden reversal of the ongoing moderation in home prices. Both new home and existing home inventories remain near record levels. Increases in foreclosures effectively add to the inventory of unsold homes, while increasing the demand for rental units.
In recent months we have witness an acceleration of rents and OER in the South and the West. Cities such as Miami, Atlanta, Houston, and the metro Washington DC area have witnessed material increases in rents over the past year.
In the West, there has been sharp acceleration of rent costs in San Francisco, and Seattle over the past year.
Against this backdrop we see little reason to anticipate any relief from the OER component of the CPI or PCE Deflators. As such, we suspect that it will be difficult to envision a near term moderation of Core inflation back into the comfort-zone. Core-Inflation is likely to be sticky in spite of an opening of the GDP Gap in the quarters ahead.
Moreover, interest rate policy that weighs on housing activity and home prices could cause the OER distortion to become more severe.
What About Non-OER Core Inflation?
The contribution to CPI from all other non-OER components of CPI was 1.4% over the 12-months ending February. This is about the average of the past 10-years excepting the 2002 to mid-2004 period when deflation was the fear.
The range in the non-OER contribution, excepting the deflation-fear period, was pretty much anchored in the 1.2% to 1.6% range. In other words, we are not seeing an inflation threat from these non-OER components.
Moreover, much of the downside deviation in the non-OER core components during the deflationary period were actually related to the drop in the funds rate feeding through to lower auto financing incentives on motor vehicles. Thus in effect the easing of policy at that time caused an outright decline in the CPI and PCE deflators for motor vehicles.
|
US
