A 3.5% utilities jump led September's rise in US industrial output

Power plants did most of the moving in September. Utilities output jumped 3.5%, according to the Federal Reserve's industrial production release published October 16, and that helped lift total US industrial production 0.7% after a 0.1% gain in August. Factories added 0.3%. The index for mining moved up 0.2%.

The quarter looks firmer than the monthly numbers did as they came in. Industrial production grew at a 2.8% annual rate in the third quarter. At 104.0% of its 2017 average, the total index stood 2.2% above September 2025. Revisions helped as well: August was first reported as flat and now shows that 0.1% gain.

For anyone planning around these numbers, the useful question is how much of the gain came from factories. On the Fed's own data, less than the headline suggests. The gap between sectors is where the information sits.

Mines and power plants carry the index

The index levels tell the rest. On the same 2017 scale, mining sits at 124.5 and utilities at 117.3. US factory output in September stood at 98.4% of its 2017 average, so whatever gains the industrial index has made above that baseline rest on mines and power plants.

The year-over-year rates rank the sectors the same way. Utilities output is up 8.7% from September 2025 and mining 2.8%, against 1.1% for manufacturing. The ranking matters because factories are by far the largest piece of the index. Design News, covering a 2015 Fed release, noted that manufacturing accounts for three-quarters of US industrial output. A smaller sector has to move a long way to set the pace, and in September utilities did.

Business equipment is the exception inside the factory story. Output of business equipment is up 6.7% on the year, the strongest of the Fed's market groups, though it slipped 0.1% in August and was flat in September at an index of 101.0. Consumer goods went the other way: a 0.7% September gain left them 0.1% below a year earlier. For a buyer of machinery and capital equipment, the annual figure points to producers who have expanded output meaningfully over the past year, with growth pausing over the last two months.

Most other market groups picked up in September after a soft August. Construction supplies rose 1.0% after a 0.7% drop, nonindustrial supplies gained 0.9% and materials added 0.7%. That left materials 2.6% above a year earlier at an index of 109.2. Materials now sit well above their 2017 average, while consumer goods (98.1) and final products (99.8) remain just under it. That split matches the industry picture and suggests output growth has been concentrated upstream.

Utilities can swing the headline in either direction, which is worth remembering before treating 0.7% as a factory signal. In September, utilities output rose 3.5% while factories added 0.3%.

Factory slack holds at 75.9%

Utilization tells a similar story with more precision. Total industrial capacity use rose to 76.9% in September from an upwardly revised 76.4%, according to Trading Economics. That is the highest since December 2023 and above the 76.5% the market expected. It is still 2.5 percentage points below the Fed's 1972 to 2025 average of 79.4%. Manufacturing edged up 0.1 point to 75.9%, against a long-run average of 78.2%.

The manufacturing rate has barely moved in a year. It was 75.8% in September 2025. Factory output grew 1.1% over that period, but the Fed estimates manufacturing capacity grew 1.0%, so the two roughly cancelled out. That explains the flat operating rate. It also means the rate on its own makes a sector that has grown look as if it is standing still.

For scale, the Fed's table puts the 1988-89 manufacturing high at 85.5% and the 2009 low at 63.4%. In a 1989 Brookings paper, economist Matthew Shapiro wrote that when the rate reached 84.6% in December 1988, some observers took it as a sign the expansion was ending and that inflation risk called for caution at the Fed. September's 75.9% is nowhere near that zone. It is also below the 77.6% Valve Magazine reported for January 2013 and the 78.4% that Design News reported for December 2014.

The stage-of-process view shows where September's movement happened. The jump came in primary and semifinished goods, the intermediate layer between raw inputs and finished products. Plants making finished goods didn't move. Relative to their own long-run norms, the primary and finished stages both sit roughly three points low by simple subtraction, while the crude stage runs near normal.

For a procurement director buying finished components or equipment, a factory sector running 2.3 points under its norm, with an operating rate flat for six months, suggests suppliers in aggregate have room to take on volume. Individual plants and product lines will vary, and an aggregate rate says nothing about a specific supplier's backlog. For buyers whose costs lean on mined inputs, mining at 87.5% is the tighter part of the chain, and the one more likely to show strain if demand picks up.

The data shows why mining is tight. Its operating rate has risen every month since April, from 85.8% to 87.5%, while the Fed estimates mining capacity shrank 0.1% over the year. Output up 2.8% on capacity that isn't growing is how a sector ends up 2.3 points above its long-run average. It also suggests relief would need new capacity or slower output, because the denominator isn't expanding on its own.

Reading the power surge with care

The utilities surge built over three months rather than arriving all at once. Output was flat in May and rose 0.3% in June, then 1.1% in July, 1.5% in August and 3.5% in September. Utilities capacity grew 3.1% over the year, the fastest of the three major industries. Output's 8.7% gain still outran it, lifting the operating rate from 70.0% a year earlier to 73.8%.

Even after a 2.3-point jump in September, that rate is 10.2 points below its historical average. On the Fed's measure, then, the power surge used up spare capacity that the sector still has plenty of.

This component can also reverse quickly. In December 2014, Design News reported, warm weather across the country cut heating demand. Utilities output fell 7.3% and total industrial production slipped 0.1%, even though manufacturing rose 0.3%. That month mirrors this September: the same 0.3% factory gain, with utilities pushing the headline the other way.

Every September figure is also marked preliminary. One month of data can move. For the 75.9% manufacturing reading, a flat rate can describe plants growing in step with demand as easily as plants sitting idle, so the capacity-growth figure belongs next to the rate whenever it's used in planning.

Academic research supports that caution. A 1999 study in the Review of Industrial Organization found that in US manufacturing, capital expansion without matching market growth, along with higher materials and capital prices, contributed to lower capacity utilization. It also found that conventional utilization measures miss changes in the economic conditions firms face. A low rate can therefore reflect investment running ahead of demand as much as weak demand.

October's numbers land November 17

FRED's histories go back a long way, which helps when building baselines. The total industrial production index starts in January 1919, manufacturing output in January 1972 and total capacity utilization in January 1967.

The Fed's next release, covering October, is scheduled for November 17. Trading Economics' models expect total utilization of 76.7% by quarter-end, a slight pullback from September. Further out, its econometric models put the rate around 77.0% in 2027 and 76.8% in 2028. Both are below the 79.4% long-run average.

Two readings will show whether this holds. If manufacturing utilization breaks out of the 75.8% to 76.0% band it has held since April, the case for slack on factory lines weakens. If utilities give back much of September's 3.5%, the month's 0.7% headline will look like a one-month power story, and factory output still won't be back to its 2017 level.