FRB Dallas Home » Research & Data »PCE »2010 »Behind the Numbers: PCE Inflation Update, July 2010

Research Publications

Behind the Numbers: PCE Inflation Update, July 2010

This update, prepared by Dallas Fed Senior Economist Jim Dolmas, provides an in-depth analysis of the latest personal consumption expenditures (PCE) inflation data. Updates will be posted monthly, following the release of the official PCE data by the Bureau of Economic Analysis. NOTE: Terms in bold are defined in the Inflation Update Glossary.

The headline PCE price index increased at a 2.9 percent annualized rate in July, breaking a string of three consecutive monthly declines. Just as those three monthly declines were driven primarily by falling energy prices, July’s increase owed much to a sharp rise in gasoline prices. The 12-month headline rate ticked up from 1.4 percent to 1.5 percent.

Core PCE posted a 1.3 percent annualized rate in July, slightly above its rate of increase over the past six months (an annualized 1.1 percent). As a consequence, the 12-month core rate held steady at the 1.4 percent rate recorded for June.

The trimmed mean PCE increased at a 0.9 percent annualized rate in July, identical to its rate of increase in June. The 12-month trimmed mean rate held steady at 1.0 percent for a third consecutive month.

Considering the latest core and trimmed mean data, the underlying trend in inflation appears—for the moment, at least—to be holding steady, albeit at a low rate.

Rebound in Energy Prices Lifts Headline Rate

As we predicted in the June inflation update, seasonally adjusted gasoline prices increased sharply in July, rising 4.4 percent on a month-to-month basis, or about 67 percent at an annualized rate. By itself, the increase in the price of gasoline accounts for about 1.5 annualized percentage points of July’s 2.9 percent annualized headline PCE rate (in the sense that excluding just gasoline from the index would lower the July rate to 1.4 percent).

Based on weekly Department of Energy gasoline price data for the month of August, we can expect a contribution from gasoline of nearly the same magnitude when PCE data for August are released. The weekly DOE data show the average retail price of a gallon of gasoline on track for an increase of about 0.5 percent (monthly rate) in August, a month where the price of gasoline typically falls by about 3.5 percent. These numbers translate into a seasonally adjusted increase of about 4 percent (or an annualized rate of about 60 percent). That’s probably more than enough to produce another positive headline inflation rate.

Other energy components—in particular, electricity (up an annualized 6.6 percent) and natural gas (up an annualized 22.2 percent)—also made noticeable positive contributions to July’s headline number.

Food Contributes Negligibly

At an annualized 1.4 percent, the number for PCE “excluding just gasoline” discussed above is awfully close to the annualized rate of 1.3 percent posted by PCE excluding food and energy in July. We shouldn’t be surprised, then, to find out that food prices were largely a nonfactor in July’s data. The price index for food as a whole—officially “food and beverages purchased for off-premises consumption”—was close to unchanged in July, up just 0.5 percent annualized.

Underlying that negligible change were nearly offsetting movements in prices at the less-processed and more-processed ends of the food spectrum. Prices for less-processed items were down 4.7 percent at an annualized rate, while prices for more-processed items—breaking with their recent behavior—were up 2.6 percent at an annualized rate. If sustained, healthier price increases for more-processed items could signal that producers of brand-name food items have begun to reclaim some pricing power.

Over the recent past, though, the behavior of prices for more-processed items has been consistent with the weak underlying inflation that we have observed—even with July’s gain, prices for more-processed items are down an annualized 0.3 percent over the past six months and are basically unchanged over the past 12 months.

Core Inflation Holds Steady

As noted above, core PCE posted a 1.3 percent annualized rate in July, up from an annualized 0.5 percent in June and identical to May’s core rate. The six-month core inflation rate actually ticked up, from an annualized 1.0 percent to 1.1 percent, while the 12-month core rate was constant at 1.4 percent.

Core goods inflation in July was basically zero—an annualized 0.04 percent—while core services inflation came in at an annualized 1.7 percent. The corresponding 12-month figures are –0.3 percent for core goods and +1.9 percent for core services.

As always, a number of items had outsized impacts on the core, in both the positive and negative directions. Jewelry and household linens had the largest negative impacts—combining to shave about a quarter of an annualized percentage point off the July core rate—while women’s and girls’ apparel, items purchased by the nonprofit sector on behalf of households, and commercial bank services each contributed more than 0.2 annualized percentage points in the positive direction.

Importantly, two of the largest core components, rent and owners’ equivalent rent (OER), continued their recent firming. Rent increased an annualized 0.9 percent for the second consecutive month, while OER increased an annualized 0.7 percent. That latter number is down from June’s 1 percent rate, but still represents a third consecutive increase, after six months of declines.

Why Is Core PCE So Much Higher than Core CPI?

For the 12 months ending in July, the core Consumer Price index was up just 1.0 percent, compared with a 1.4 percent rate for the core PCE index. With both core rates very low, a difference of 0.4 percentage points is a substantial one. What explains this gap between the two 12-month core rates?

The CPI and PCE differ in a number of ways, such as the frequency with which items’ weights are updated (monthly for the PCE, biannually for the CPI) and the scope of the index’s coverage (PCE, for example, includes prices for items purchased by the nonprofit sector on behalf of households, while CPI only looks at households’ direct expenses ).

While all the differences in the construction of the two indexes play some role in explaining the indexes’ recent divergence, two differences are likely suspects for having played more significant roles. First, compared with core PCE, core CPI places a much larger weight on shelter costs, which have decreased slightly over the past 12 months. And second, core PCE places a much larger weight than core CPI does on medical care costs, which have risen more rapidly over the past 12 months. The differences in weighting are large—OER and rent of primary residence add up to about 17 percent of core PCE versus about 37 percent of core CPI, while medical care adds up to about 22 percent of PCE versus about 8 percent of core CPI.

Does this difference in the weighting of shelter and medical care explain the 0.4 percentage point gap between the two 12-month core measures? It turns out that it more than explains the gap: If we adjust the weights in core PCE so that shelter and medical care have the same relative importance as they do in core CPI, the 12-month core PCE rate falls by 0.5 percentage points, from 1.4 percent to 0.9 percent.

Evidence from the Trimmed Mean

Core PCE can, as we have often seen, be buffeted by large price swings in a small number of items, making it difficult to infer underlying inflation trends. The recent behavior of the trimmed mean, though, tends to corroborate the “stable for the moment” impression one gets from core PCE. One-month trimmed mean inflation rates for each of the past three months (annualized rates of 0.8 percent, 0.9 percent and 0.9 percent) have exceeded the corresponding six-month rates for each of those months (0.7 percent in each month).

The past two one-month readings are just below the trimmed mean’s 12-month rate, constant at 1.0 percent for the past three months.

A Decline in the Number of Falling-Price Items

July saw a small drop in the number of falling-price components among the 178 components that potentially go into the trimmed mean. Of those 178 components, 66 registered declines in price, down from 73 in June.

Nevertheless, the fraction of components experiencing price declines remains elevated compared with our average experience over the past 30 years.

—Jim Dolmas
August 30, 2010


Federal Reserve Bank of Dallas Seal
Federal Reserve Bank of Dallas

2200 N. Pearl St., Dallas, Texas 75201 | 214.922.6000 or 800.333.4460
Disclaimer / Privacy Policy

Federal Reserve Centennial