Free weekday email: what moved overnight, and what it changes.
Get the Daily Briefor the In-Depth EditionOctober retail sales rose 0.2 percent, a change inside the survey's own error band, while industrial production fell 0.4 percent almost entirely because utilities dropped 4.3 percent. The composition of both releases, not the headlines, shows gasoline prices carrying the spending total and factories still adding output.
Data as of November 17, 2026.
Two U.S. releases landed within hours of each other on November 17 and appear to point in opposite directions. Retail and food services sales were $748.2 billion in October, up 0.2 percent on the month and 6.8 percent on the year. Industrial production fell 0.4 percent. Neither headline describes what happened to demand: the retail change sits inside its own confidence interval, and the production decline is almost entirely a single weather-sensitive sector. The composition of each release does describe it, and the split inside both is the story.
Advance estimates of U.S. retail and food services sales for October were $748.2 billion, up 0.2 percent from September and up 6.8 percent from October 2025, the Census Bureau reported. The advance estimate carries a margin of error of plus or minus 0.4 percent at the 90 percent confidence level, so the monthly change is not statistically distinguishable from zero. Sales across August through October were up 6.4 percent from the same three months a year earlier, and the August to September change was unrevised at 0.8 percent.
The year-over-year figure carries more information, and at 6.8 percent nominal spending it is growing roughly three times as fast as the 2.1 percent industrial output gain reported in the same week, a wedge consistent with prices still rising. The retail series is not price-adjusted, which is why the October CPI release matters as its companion: nominal spending can hold up while the volume underneath it stalls.
The advance table breaks sales out by kind of business, and the arithmetic is unusually clean this month. Total retail and food services sales rose $1.58 billion from September. Gross gains across categories came to $4.48 billion and gross declines to $2.89 billion.
The three largest gainers were nonstore retailers at $1.57 billion, gasoline stations at $1.03 billion, and miscellaneous store retailers at $0.74 billion. Those three alone are 2.1 times the net increase. Health and personal care stores added $0.59 billion, food services and drinking places $0.23 billion, building material and garden stores $0.21 billion, and general merchandise stores $0.11 billion.
Motor vehicle and parts dealers subtracted $2.23 billion in a single category, more than the entire net gain. Furniture stores fell $0.25 billion, clothing stores $0.20 billion, food and beverage stores $0.15 billion and sporting goods, hobby and book stores $0.06 billion.
A month in which the total barely moves can therefore contain a $2.2 billion drop in one category and a $1.6 billion rise in another, and the composition carries more signal than the sum. Sales excluding motor vehicles rose 0.6 percent. Excluding both vehicles and gasoline, they rose 0.5 percent.
Gasoline station sales were $68.6 billion in October, up 30.3 percent from $52.6 billion a year earlier on the same seasonally adjusted basis. No other category in the table comes close. Nonstore retailers, the strongest genuine growth channel, rose 11.8 percent. Food services rose 7.1 percent.
A gas station's takings are price multiplied by gallons, and stations sell close to the same volume week to week, so a 30 percent increase is almost entirely the price of fuel. Gasoline accounted for 9.2 percent of October's $748.2 billion. When fuel prices rise, retail sales rise and households have less left over for everything else, which is why the two series move together in the opposite direction from household comfort.
The gap between the two measures is the cleanest read on the price effect in the release. Sales excluding gasoline stations rose 4.9 percent on the year, against 6.8 percent for the total. The 1.9 point difference is what fuel prices added to the headline without adding to the volume of goods and services sold.
The Federal Reserve's G.17 release reports total industrial production down 0.4 percent in October, after a 0.7 percent gain in September. Underneath that headline, manufacturing output rose 0.3 percent. Mining fell 0.8 percent and utilities dropped 4.3 percent.
Utilities are roughly a tenth of the production index, so a 4.3 percent drop there subtracts close to 0.4 points on its own. Manufacturing, at about three quarters of the index, added back around 0.2 points and mining subtracted about 0.1. The headline decline is essentially the utility move plus rounding, which is why the Federal Reserve's own summary leads with utilities and treats manufacturing as the detail. Utility output had gained 4.4 percent in September, and October's 4.3 percent fall more than reverses it: the index sits at 111.7, just under August's 111.8.
Manufacturing excluding motor vehicles and parts edged down 0.1 percent, so vehicle assembly was the small positive inside a soft factory month. That is the opposite of the retail report, where motor vehicle dealers were the largest single drag at minus $2.23 billion. The two releases can disagree about the same sector because they measure different things, dealer revenue in one and factory volume in the other. Dealer revenue fell while assembly rose, a combination that builds vehicles in stock unless retail demand absorbs them.
Total industrial production was 103.4 percent of its 2017 average in October, 2.1 percent above the year-earlier level. The Federal Reserve also revised the September level down to 103.8 from 104.0 while leaving the monthly gain at 0.7 percent, so the level is lower even though the direction is unchanged.
Capacity utilization fell to 76.4 percent in October, which the Federal Reserve describes as 3.0 percentage points below its long-run 1972 to 2025 average. An industrial base running 3 points below its average utilization, with output 2.1 percent higher than a year earlier, is not short of capacity. Production growth is coming from existing plant rather than from new plant being strained. It also means the price pressure visible in the retail data is not originating on the factory floor. The euro area's September production figures show a similar pattern from the other direction: a small monthly gain that still leaves output below last year.
The Federal Reserve meets December 8 and 9. The case for another move, argued publicly by Christopher Schmid on the ground that inflation remains a demand problem, depends on the real policy rate being low: with PCE inflation at 3.6 percent against a 3.75 to 4.00 percent funds range, the ex-post real rate is about 0.28 percentage points. That argument is set out in Schmid's demand-side case and in the measure-mismatch analysis that shows how the choice of inflation index decides whether December tightening looks necessary or premature.
October's pair of releases does not settle it. Nominal spending grew 6.8 percent while industrial output grew 2.1 percent, a wedge consistent with inflation that has not returned to target. But a monthly retail print of 0.2 percent inside its own error band, and capacity utilization 3 points below its average, are not the marks of demand accelerating into year end.
The labour side is already in. The October jobs report showed payroll growth of 165,000 with wages still rising, and the mortgage spread work explains why a move at the short end reaches a household payment slowly and unevenly. The repricing mechanics piece traces how a December change travels along the curve, and the four-bank comparison puts the real-rate argument in the international frame where other central banks are making the same calculation.
Census publishes the advance retail estimate for each month about four weeks after the month ends, so the October figures arrived on November 17 and the November figures will not exist when the committee votes. The decision will rest on October and on the revisions to it.
Two numbers carry the most weight from here. First, the November gasoline station line: if station sales stop climbing, the 6.8 percent year-over-year headline loses roughly a fifth of its support and the underlying consumption trend becomes visible. Second, utilities output in November, because a 4.3 percent drop tied to mild weather reverses, and the industrial production headline will follow it back up without any change in factory demand.
Between those two, the week in markets has already shown how quickly a single month of soft goods data gets read as a growth scare by equity investors while the bond market reads the same print as persistent inflation. The November 17 releases give both sides something: a retail total that is flat in real terms, and a production index whose weakness is concentrated in the one sector that moves with the weather.
Originally published by Global Financial Info. https://globalfinancialinfo.info/article/retail-sales-0-2-percent-industrial-output-fall-october-split-november-17-2026
Licensed CC BY-NC 4.0Free to republish with credit and a link. Commercial use requires permission.
Global Financial Info. (2026, November 18). Retail Sales Rose 0.2 Percent as Output Fell 0.4 Percent: The Split Inside Both October Releases. Global Financial Info. https://globalfinancialinfo.info/article/retail-sales-0-2-percent-industrial-output-fall-october-split-november-17-2026
Free for noncommercial use under Licensed CC BY-NC 4.0. Commercial licensing: partners@globalfinancialinfo.info