PrintPrint EmailEmail ShareShare CiteCite


How High Is U.S. Inflation?

Author: Roger M. Kubarych
January 25, 2005
Council on Foreign Relations


The latest data on production and spending confirm that the US economy is entering 2005 with considerable momentum. The Fed’s industrial production index was up a solid 0. 8% in December and is now 4. 4% higher than a year ago. Retail sales for December also advanced a remarkable 1. 2%, pulled up by a tremendous 4. 3% surge in sales of autos and parts. Despite retailer complaints early in the month that the Christmas sales season was disappointing, non-auto sales managed to increase by 0. 3%. Evidently, last-minute shoppers before the holiday weekend and determined bargain hunters in the last week of the month pulled up non-auto sales. Overall retail sales are now running 8. 7% above year-ago levels, an unequivocally strong performance. Housing starts and new building permits in December also rebounded sharply after a brief decline caused mainly by bad weather and the aftermath of last fall’s hurricanes. There may be latent weaknesses in the US economy but they are hardly apparent in the data.

Moreover, anecdotal comments to Fed researchers by local business contacts reinforce the strong tone. The Beige Book compendium of business comments reported that eleven of the twelve Fed district banks characterized economic activity as expanding, with special strength in the Atlanta, New York and Richmond districts. (Richmond includes Washington D. C. ) Only the Cleveland district, an area of aging industrial facilities, injected a more cautious note, describing business activity as mixed.

New statistical and anecdotal information about the inflation situation also came out during the past week. Both the PPI and the CPI for December reflected the sharp, but evidently temporary, drop in energy prices that took place in the past couple months. The finished goods PPI fell by 0. 7% overall, while the core PPI for finished goods excluding food and energy prices ticked up only a tenth of a percent. From a year ago, the finished goods PPI is now up by 4. 1%, while the core finished goods PPI is up 2. 2%. The CPI results were broadly similar. There was a modest 0. 1% decline in December largely because of a big 1. 8% drop in energy costs. That leaves the overall CPI up 3. 3% from a year ago. By comparison, the core CPI was up another 0. 2%, for a year-over-year rise of 2. 2%. All of these indexes show a sharp acceleration from the rate of inflation in 2003, which was almost a percentage point lower. So inflation accelerated during 2004, but tapered off at the end of last year.

Source: Bloomberg, BEA

There are other ways of looking at inflation developments that are not so benign. In particular the rise in average home prices, as measured by the respected survey conducted by the Office of Federal Housing Enterprise Ownership (OFHEO), over the twelve months through October 2004 is a staggering 13%. The annualized rate of increase in just the third quarter of 2004 alone is a blistering 18%. Nothing comes close to signifying the reemergence of an inflationary psychology in the US than the run-up in housing prices, fueled by abundant, cheap mortgage financing and homebuyers who are shunning other assets, especially equities and equity mutual funds, because of perceived risks. Presumably they see little risk in the housing market and are rushing out to buy new and existing homes before prices go up further. That is a classic manifestation of a growing inflationary mentality, and the Federal Reserve, in the minutes of the December FOMC policy meeting, called attention to speculative behavior in the housing market.

However, the methodology by which the Bureau of Labor Statistics computes the consumer price index does not show anything remotely resembling this frantic behavior. Home prices do not appear directly at all in the two main consumer price indexes in the US, the CPI and the separate personal consumption expenditure price index (PCEPI) calculated by the Bureau of Economic Analysis as part of the exercise for determining nominal and real GDP. They use a concept called “owners’ equivalent rent of primary residence.” They survey the small number of homes that are rented and then make the assumption that homeowners pay themselves a rent equivalent to what is happening in this narrow rental market for single-family homes. This category carries the single-most important weight in the calculation of the CPI (although it has a much lesser weight in the PCEPI). And it is up only a modest 2.3% over the past twelve months, a tiny fraction of home price rises in the marketplace. While most economists support the methodology behind the calculation, the huge disparity raises serious questions about whether the consumer price index is understating the inflationary consequences of rising home prices.

Source: Bloomberg,

Recently, the Federal Reserve has indirectly blessed the alternative index for measuring inflation at the consumer level by shifting the commentary and the published forecasts to the PCEPI from the CPI. The PCEPI usually shows lower inflation than the CPI. Over the past forty years, it has averaged about 0.7 percentage point a year lower. But in the past few years, changes in the way both the CPI and the PCEPI are calculated has narrowed the divergence.

Source: Bloomberg

The trouble with the PCEPI is that it gives results that don’t square with common sense – nor with alternative sources of information. In particular, it places an enormous weight on the price of medical services. They account for a whopping 20% of the total index and 25% of the core index. By contrast, the weight for medical care in the CPI is just a little over 6%. That means that the validity of the PCEPI as an indicator of consumer inflation depends to a very large extent on the way medical care costs are estimated. And the unique method that the BEA uses to estimate those costs sometimes gives totally bizarre results. In particular, in the past year it has estimated that medical care prices have actually decelerated sharply, from a rate of 4% per annum to under 2. 5% per annum. This differs from the CPI’s different methodology, which shows medical services costs going up by 4.4% over the 12 months of 2004. Most of the difference between the CPI and PCEPI can be attributed to the divergence in the two price series in this important category.

Source: Bloomberg, BEA,

There is one clue that the CPI medical care price index is more credible than the PCEPI. That is found in the quarterly employment cost index, which splits out wages and other forms of compensation. Costs of benefits, mainly health care insurance, have been rising at a rate almost double that of wage increases, close to 6% per annum. That is far above the estimate of medical price increases that appears in the PCEPI.

Source: Bloomberg

This technical dispute might not amount to much except for the important fact that the Federal Reserve has adopted the PCEPI as its preferred measure of consumer inflation. But it may not be a more reliable indicator than the more familiar CPI, notwithstanding the statistical problems associated with that index. But there are some signs that methodological changes over the next few months will lead to an upward revision of PECEPI.

What is the Fed hearing from business contacts about US inflation? According to the Beige Book, inflationary pressures remained largely in check in December and early January. “While many manufacturers and builders continued to report small increases in input costs, price increases for final goods and services were generally modest. ” And many retailers told the Fed that competition was keeping them from raising prices to the consumer. This restrained inflation picture, if it holds up, will give everyone time to debate which price index is best. But until that controversy is resolved, the scattered signs of rising inflationary expectations are beginning to multiply.

More on This Topic