On Thursday morning, the Labor Department delivered a number that would have, in almost any other context, brightened the mood of central bankers: core inflation fell to 2.4% in August, its lowest reading since March 2021. By the close of trading, futures markets had responded by pricing a Federal Reserve rate hike at nearly 89 cents on the dollar. Something unusual is going on — and the gap between those two data points is the story.
The full August consumer price index, released September 11, showed headline inflation holding steady at 3.4% year over year, with the monthly gauge rising a seasonally adjusted 0.4%. Core CPI — which strips out food and energy to capture underlying price trends — ticked down to 2.4% annually, the sixth consecutive month of year-over-year deceleration and the lowest reading in more than five years. Shelter costs, which had been among the most stubborn components of the basket, slowed to 3.0% year over year, down from 3.2% in July. Core goods posted a modest 0.7% annual gain. On paper, this looks like a report the Federal Reserve should welcome.
The problem is that the bond market, and increasingly the Federal Reserve, has learned to read beyond the headline. Within hours of the release, the 10-year Treasury yield surged 18 basis points (a unit equal to one-hundredth of a percentage point) to 4.96%, while the two-year — which anchors most tightly to near-term Fed expectations — jumped 26 basis points to 4.63%. The yield curve compressed, and hike odds on the CME's FedWatch Tool (a market-derived probability gauge based on fed funds futures) rose above 85%. The benign core reading had been eclipsed by three inconvenient facts: energy is re-accelerating, the Fed's actual preferred inflation gauge paints a considerably more worrying picture, and the labor market is too healthy to force the committee's hand toward a pause.
The Measure That Actually Matters
The Federal Reserve does not officially target the consumer price index. Its formal inflation mandate is anchored to the personal consumption expenditures price index, or PCE, a separate measure produced by the Bureau of Economic Analysis. The two gauges differ in methodology, weights, and scope — and right now they are telling divergent stories. Core PCE for July, the most recent available reading, came in at 3.3% year over year, roughly 90 basis points above core CPI's August reading of 2.4%. That gap has persisted since late 2025, when core PCE overtook core CPI for the first time since 2021.
The reasons for the divergence are structural. PCE assigns considerably greater weight to healthcare services and other categories where price growth has proven stickier. It captures a broader range of consumer spending, including out-of-pocket medical expenses poorly represented in CPI's household survey basket. PCE also uses chain weighting — a method that adjusts for shifts in what consumers actually buy rather than a fixed market basket — making it more sensitive to substitution effects in services. The result is a measure that the Fed views as a more accurate read on the underlying price level. At 3.3%, core PCE is still running 65% above the 2% target. That number, not the CPI figure that dominated headlines Thursday morning, is what policymakers will have in front of them next week.
The Energy Wildcard
Even setting PCE aside, the trajectory of energy prices is creating fresh anxiety about the second half of 2026. Gasoline prices rose 3.9% in August alone and are now up 27.4% over the past twelve months. Fuel oil has surged 52% year over year. The energy index as a whole accounted for more than a third of the total monthly CPI increase. In isolation, energy volatility is something the Fed typically looks through when setting policy — transitory swings in oil don't change the underlying structural path of services inflation. But sustained energy elevation has a more insidious effect: it feeds into production costs, transportation, and eventually the prices of the services that the Fed is most focused on controlling. The pipeline from gasoline to core services runs on a three-to-six month lag, and it is filling again.
| Indicator | Aug 2026 | Jul 2026 | Change |
|---|---|---|---|
| Headline CPI (YoY) | 3.4% | 3.4% | — |
| Core CPI (YoY) | 2.4% | 2.5% | −0.1pp |
| Shelter (YoY) | 3.0% | 3.2% | −0.2pp |
| Energy (YoY) | +16.3% | +14.7% | +1.6pp |
| Gasoline (YoY) | +27.4% | +24% | ↑ |
| Core Services (YoY) | 3.0% | 3.0% | — |
| Core PCE (YoY, Jul) | 3.3% | Flat 2 months | |
A Labor Market That Won't Cooperate
If cooling core inflation alone were the full picture, the Fed might find reason to hold at its September meeting. But the August employment report, released six days before the CPI print, removed that option. Nonfarm payrolls — the monthly count of jobs added across the economy — expanded by 162,000, more than triple the 53,000 consensus estimate. The unemployment rate held at 4.1%, and average hourly earnings rose 3.1% year over year. This is not the fraying labor market that would typically give a central bank cover to pause its tightening campaign.
Fed Chair Kevin Warsh set the rhetorical tone at the Jackson Hole symposium on August 28, declaring that inflation "remains stubbornly above" target and that the Fed "has more work to do." That language, combined with the three dissents at the July 29 meeting — all three Fed officials voted for an immediate hike rather than the hold the committee chose — signals a committee already tilted toward action. The August CPI report gave policymakers neither a compelling reason to hold nor any data that would accelerate the pace beyond the expected 25-basis-point increment. It confirmed a meeting that, in all likelihood, was already decided.
What to Watch Next
The September 16–17 FOMC meeting will produce a new set of quarterly economic projections alongside the rate decision, including the dot plot — a chart displaying each policymaker's anonymous forecast for the fed funds rate through 2028. If the median 2026 dot is revised upward from the June projection, it signals the committee sees additional hikes on the table before year-end. Retail sales for August, due Tuesday morning before the FOMC opens, will provide the final major data point before the decision is announced Wednesday afternoon.
The most consequential number for the next phase of the rate debate will arrive two weeks later. Core PCE for August is due around September 26. If it follows core CPI lower — closing any portion of the 90-basis-point gap between the two measures — it would suggest the underlying disinflationary trend is real and potentially restrain the pace of further hikes. If core PCE holds at 3.3% or re-accelerates, the Fed has its next warrant. This week, the September hike is all but decided. The market is already pricing what comes after it.
The August CPI report contained genuinely encouraging news: core inflation is at its lowest since early 2021 and shelter costs are finally moderating. But the Federal Reserve targets PCE, not CPI, and its preferred gauge remains at 3.3% — flat for two consecutive months. With energy prices re-accelerating, the labor market at nearly triple the expected pace, and the committee already signaling its intentions, a 25-basis-point hike on September 17 is all but priced. The dot plot will answer the harder question: whether one more is the end of the cycle or the beginning of a new leg.
