Core services inflation holds at 3.35% after the BEA's PCE method change

On the revised series, core services prices rose 3.35% from a year earlier in July and 3.35% again in August ^[4]. Measured on one consistent method, the rate did not slow between the two months ^[2]. Core services are more than 60% of consumer spending ^[5]. At 3.35% they run 1.35 percentage points above the Fed's 2% target ^[1][1]. Wolf Richter, who reported the revisions, wrote that July's 34-basis-point cut came from "these methodological changes and other revisions" ^[4]. His account does not split the two.

Most of the other gauges rose in August. Core PCE went up 3 basis points from the revised July figure, to 3.01% ^[9][6], and the all-items index rose to 3.42% from a revised 3.36% ^[11]. Core goods were the exception, falling to 3.07% from a revised 3.61% in July, a 54-basis-point drop ^[7][7]. Core PCE has been above target since March 2021, and its closest approach on the revised history was 2.6% in April 2025 ^[10].

The case for taking the lower numbers at face value has senior backers. Fed Governor Christopher Waller said that "nonmarket services prices have always been an issue for me, since they are imputed and not actual price changes," and that "my take is that underlying inflation is doing better than the core numbers suggest" ^[13]. Richter lists the changed categories as portfolio management and investment advice, legal services, and computer software and accessories ^[3]. American Banker described the change as a new way of estimating nonmarket costs such as fees charged for financial services and insurance products ^[12]. Kevin Hassett, director of the White House National Economic Council, argued that if inflation over the last three months is running around 2%, and interest rates since the early 19th century have been about the inflation rate plus two, then there is not a lot of room for rates to go up from here ^[14].

The two monthly core readings in the release support part of that. July's revised increase and August's 0.25% compound to an annualized pace of about 2.2% ^[8][3]. August alone annualizes to about 3.0% ^[4]. The near-2% short-run figure depends on the July month the revision halved ^[8].

If the new method prices those fees more accurately, the old 3.69% overstated inflation and the revision corrected an error ^[4]. If monthly core prints settle near July's revised level, Hassett's short-run figure is right and the year-over-year rates follow it down over the next year. The third possibility is a one-time level shift, with year-over-year rates rising from a lower base as they did in August ^[9][11]. I think the evidence so far fits the third best, though the drop in core goods cuts against it ^[7]. A run of monthly core prints at or below about 0.165%, the pace that compounds to 2% a year, would prove that view wrong ^[5].

New York Fed President John Williams spoke in Buffalo on Tuesday, the day before the revisions came out ^[16][18]. "If the economy evolves in a manner broadly consistent with my forecast, one further upward adjustment of the federal funds target range may be appropriate late this year to support a timelier return of inflation to target," he said ^[16].