Market Data

What the Manufacturing Industrial Production Index Actually Measures, and Why It Sits at 98.64

A plain-English guide to the Fed's factory-output gauge: how the 2017=100 base works, what feeds the number, and what a reading below 100 tells you about real production.

The Manufacturing Industrial Production index measures the real, physical output of U.S. factories against a 2017 base of 100, and its latest reading of 98.64, as of May 2026, per the Federal Reserve's G.17 release, means factory output is running about 1.4% below its 2017 average. Unlike dollar-based series, it strips out price changes entirely: it is a volume gauge, up about 1.4% from a year ago and climbing.

How the 2017=100 base works

The Fed sets average monthly factory output in 2017 equal to 100 and expresses every other month relative to it. A reading of 105 means factories are producing 5% more physical stuff than they did in an average 2017 month; 95 means 5% less. The concrete translation: a plant that shipped 1,000 units a day in 2017 and tracked the index exactly would be shipping about 986 a day at today's 98.64. The base year is an accounting convention, not a target, but because 2017 was a normal, mid-expansion year, the 100 line has become a useful shorthand for "pre-trade-war, pre-pandemic normal."

What feeds the number

The Fed builds the index from physical product data where it exists, tons of steel, barrels refined, vehicles assembled, and from production-worker hours and electricity usage where it does not, weighted by each industry's value added. Manufacturing is the dominant slice of the broader industrial production index, which adds mining and utilities. Durables and nondurables each contribute roughly half of manufacturing, so a strong quarter in autos or machinery can be masked by weakness in chemicals or food, and vice versa. The monthly release also arrives with meaningful revisions: the first print draws on partial data, so treat any single month as provisional until two revisions have passed.

The index also has a sibling worth knowing: capacity utilization, which is simply this output measure divided by the Fed's estimate of what factories could produce at full sustainable rates. The pair answers different questions, the production index says how much is being made; utilization says how close the system is to its ceiling. A rising index with flat utilization means capacity is being added as fast as output grows; a rising index with surging utilization means the slack is running out and lead times are about to prove it. Sector detail matters just as much: the G.17 release breaks manufacturing into dozens of industry groups, and the aggregate routinely hides a boom in one (aerospace, semiconductors) offsetting a slump in another (chemicals, furniture). A plant manager benchmarking against the headline should pull the index for their own three-digit industry code first, the national number is the context, not the comparison.

Manufacturing industrial production, index 2017=100, May 2026: 98.64. Archived range: 96.99 in Dec 2025 to 98.64 in May 2026. The latest reading sits 100% of the way up that range.

Reading a print near the baseline

Today's 98.64 sits 1.4 points below the 2017 base, and the year-over-year move, up about 1.4% from a year ago, matters more than the level. Because the index is real, that change is volume: it either is or is not being matched by your plant's own unit output. Operations leaders use it two ways. As a benchmark: if the index for your sector is climbing while your volumes do the opposite, the gap is company-specific and worth diagnosing. As a deflator check: pair it with nominal series like manufacturers' new orders to separate price from volume in your demand picture.

It is a volume gauge, not a dollar gauge, when this index moves, physical output moved.

Use the capacity planning calculator to translate a production index move into the machine hours and shifts your plant actually needs. Benchmark your own output

Published 2026-07-13.