Is American Manufacturing Actually Declining?

Is American Manufacturing Actually Declining?

LEARN / REINDUSTRIALIZATION

Is American Manufacturing Actually Declining?

Employment fell and that is not disputed. Whether it constitutes decline depends on which measure you accept as the test, and the three serious answers disagree about what should be measured rather than about the underlying facts.

Key points

What is not disputed

Manufacturing payroll employment peaked in June 1979 and stood at 12.6 million in July 2026, roughly 35 percent below that peak. Every participant in this argument accepts those figures. The disagreement begins immediately afterward, at the question of what a fall in factory employment demonstrates about a country’s ability to produce.

Position one: employment fell because productivity rose

The strongest version of this position runs as follows. Manufacturing employment fell for the same reason agricultural employment fell a century earlier, because output per worker rose faster than demand for the output. Fewer people are required to produce the same goods, labor moves to sectors where it is more valuable, and the economy is better off. On this reading, treating falling factory employment as decline confuses a productivity gain with a loss, and a policy aimed at restoring 1979 employment would be a policy aimed at making workers less productive.

The argument is coherent and it carries a load-bearing empirical claim: that output continued rising while employment fell. That claim is checkable.

Data98.7
Industrial Production: Manufacturing, June 2026, index 2017 = 100, seasonally adjusted. Real manufacturing output sat 1.3 percent below its 2017 average. Source: Board of Governors of the Federal Reserve System, G.17 release, series IPMAN.

Series and provenance
Agency
Board of Governors of the Federal Reserve System
Program
G.17 Industrial Production and Capacity Utilization
Series
IPMAN
Measure
Industrial Production: Manufacturing (NAICS)
Units
Index, 2017 = 100
Adjustment
Seasonally adjusted
Period
January 1972 to June 2026
Latest
98.6995, June 2026
Source tier
Primary
Retrieved
August 9, 2026

An index against a base year, not a level. Counts establishments located in the United States regardless of ownership; excludes U.S. territories. Rebased periodically, so readings are not comparable across rebasings without the underlying series.

Open the source

Over the long horizon the productivity argument holds comfortably, because real manufacturing output today is far above what it was in the 1970s while employment is far below. Over the recent horizon it does not hold at all. Between 2017 and mid-2026 the output index went nowhere, which means the period cannot be described as one in which fewer workers produced more. The position is therefore correct about the twentieth century and unsupported about the last decade, and which of those an argument is really about is rarely stated.

Measurement noteThe industrial production index is rebased periodically, most recently to 2017 = 100. A reading of 98.7 describes output relative to the 2017 average and says nothing about the level in 1979 or 1999. Comparisons across rebasings require the underlying series rather than the headline number, and much published commentary uses the headline number as though it were a level.

Position two: aggregates conceal the loss of specific capabilities

The second position accepts the output figures and argues that they answer the wrong question. An aggregate index sums across everything a country makes, so a rise in the production of one category can offset the complete disappearance of another. A country that stops producing a class of components entirely, while expanding production of something else by an equivalent value, records no change in the aggregate and has lost the ability to make the first thing.

This is the position that treats capability as the unit of analysis. It asks whether a specific item can be produced at a required rate within a required time, which no aggregate series was designed to answer. Its weakness is that capability inventories are expensive, program-specific, and largely unpublished, so the position is harder to demonstrate than to assert.

Capacity utilization adds a wrinkle that both of the first two positions have to accommodate. The Federal Reserve reported manufacturing capacity utilization of 77.5 percent in July 2025, about two percentage points below its long-run average since 1972. Installed capacity is not fully used, which complicates any account that treats the constraint as purely physical.

Position three: the measures are territorial

The third position observes that the standard series draw a border and then measure inside it. The Federal Reserve states the rule explicitly: the industrial production index measures the real output of all relevant establishments located in the United States, regardless of their ownership, and excludes those located in U.S. territories.

Two consequences follow. A plant in Kentucky owned by a Japanese company counts fully toward American industrial production. A plant in Mexico owned by an American company counts not at all. If the question is what the country can produce on its own soil during a disruption, the territorial measure is the right one. If the question is what American firms command in productive assets worldwide, it understates substantially, and the Bureau of Economic Analysis publishes a separate account of multinational activity for that purpose.

Institute analysisThe three positions are not competing hypotheses about one fact. They are three different questions wearing the same words, and each has a defensible answer that contradicts the others only because the questions differ. The Institute’s position is that the capability question is the decision-relevant one for national capacity: a government that knows aggregate output but cannot say whether a specific component can be produced at rate has measured something accurately and learned nothing it can act on.

Common misconceptions

That one series can settle it. Employment, output, capacity utilization and multinational activity each answer a different question, and a participant citing one series against another is usually changing the subject rather than winning the argument.

That rising output disproves decline. It disproves the claim that total production fell. It says nothing about whether particular capabilities were lost, because an aggregate cannot distinguish composition from level.

That the productivity explanation is timeless. It describes the long postwar record well and does not describe the period since 2017, when the output index was flat. An argument that relies on it should say which decades it means.

What the evidence says, and where it is contested

ContestedThe genuine dispute is over the standard of evidence, not the data. Position one requires only aggregate series and can be checked by anyone. Position two requires capability inventories that are program-specific and mostly unpublished, which makes it harder to falsify and correspondingly harder to verify. Position three is a definitional observation with which everyone agrees once stated, and which changes the answer only if the question concerns sovereign productive capacity rather than national economic performance. A reader deciding among them is choosing a standard of proof, and should say so.

Related Institute research

When Did American Deindustrialization Begin?
Three series, three starting dates, and what each can and cannot date.

What Is the U.S. Industrial Base?
Why capability and output measure different things.

The Production Gap That’s Costing America the Technology Race
The distance between design and manufacture.

Sources

Reference entry maintained by the Institute for American Manufacturing & Technology. Figures are drawn from primary sources and cited above. Where the Institute states a position rather than a finding, it is marked as such.

Private Professional Network

Institute for American Manufacturing & Technology

Join the IAMT Network. Connect with researchers, policymakers, and industry leaders working on American manufacturing, energy, and technology policy.

Apply for Membership →