What Does Reindustrialization Cost?

What Does Reindustrialization Cost?

LEARN / REINDUSTRIALIZATION

What Does Reindustrialization Cost?

Cost has three components, and the one that dominates is time. The published spending figures are nominal, gross, and measured at a different stage of the process from the announcements they are usually compared against.

Key points

The three components

Capital cost is what appears in announcements: land, structures, equipment and installation. Where the equipment is a long-lead item the cost is partly a date rather than a price, which Long-Lead Equipment and Transformers quantifies for grid equipment. Operating cost is what determines whether the plant survives after the incentives lapse, and it covers energy, labor, materials and maintenance across the asset life. Time cost is the least discussed and usually the largest, because a plant that reaches rate three years behind schedule has forgone three years of output and three years of the market position it was built to hold.

Analyses that report only capital cost are describing the component that is easiest to observe and easiest to subsidize. A construction grant reduces the first and leaves the second and third untouched, which is why plants built with public support can still close when operating economics turn.

What the construction data shows

The Census Bureau publishes monthly value of construction put in place, including a manufacturing category. That series is the closest thing to a real-time measure of industrial buildout, because it records work actually done rather than work announced.

Data$223.1B → $195.3B
Private manufacturing construction spending, seasonally adjusted annual rate, June 2025 to January 2026. A decline of roughly 12 percent across seven months. Source: U.S. Census Bureau, series PRMFGCONS.

Series and provenance
Agency
U.S. Census Bureau
Program
Construction Spending, Value of Construction Put in Place
Series
PRMFGCONS
Measure
Total Private Construction Spending: Manufacturing
Units
Millions of dollars, annual rate
Adjustment
Seasonally adjusted; NOT adjusted for price changes
Period
January 1993 to January 2026
Latest
$195,309 million, January 2026
Source tier
Primary
Retrieved
August 9, 2026

Census states the data are adjusted for seasonality but not for price changes. During periods of construction inflation the real change is larger than the nominal change in both directions.

Open the source

The direction matters more than any single month. Manufacturing construction rose steeply through the early 2020s and has been falling since mid-2025, which means the buildout that dominated recent coverage is now contracting on the measure that tracks completed work.

Measurement noteCensus states plainly that its construction data are at an annual rate, adjusted for seasonality but not for price changes. Because industrial construction costs rose substantially over the same period, a nominal series that is flat describes real activity that fell, and a nominal series that fell 12 percent describes a larger real decline. The BLS producer price index for new industrial building construction and the BEA price index for manufacturing structures exist to make that adjustment, and most commentary does not apply either.

Announced, obligated, outlaid

Three distinct quantities circulate in coverage of industrial policy, and conflating them is the most common error in reporting on program scale. An announcement is a statement of intent by a company or an agency. An obligation is a binding commitment of federal funds. An outlay is money that has actually left the Treasury.

The three diverge widely and in a predictable direction. Announcements exceed obligations because plans change and conditions are not met. Obligations exceed outlays because funds are drawn against milestones over years. A program described by its announcement total and assessed by its outlay total will appear to be failing when it is merely early, and a program described by announcements alone will appear larger than anything that has happened.

Institute analysisThe useful discipline is to state which of the three a figure represents every time one is used, and to compare like with like across years. A comparison between one year measured by announcements and another measured by outlays is not a comparison.

Why cost per job is the wrong metric

Public discussion frequently divides a program total by expected employment and reports the result as cost per job. The calculation is arithmetically simple and conceptually confused, because the purpose of building productive capacity is capacity rather than headcount.

A highly automated facility producing a component the country cannot otherwise obtain will show a terrible cost per job and may be the more valuable investment. A labor-intensive plant assembling imported components will show a better figure and add little to what the country can make. The metric systematically ranks the second above the first, which is the opposite of what a capacity policy is trying to achieve.

Common misconceptions

That announcement totals measure investment. They measure stated intent. Construction put in place measures work performed, and the two series diverge by a wide margin.

That a nominal decline is the whole decline. Construction spending is not price adjusted. During periods of construction inflation the real change is larger than the nominal change in both directions.

That subsidised construction settles operating economics. A grant reduces capital cost once. Energy, labor and materials recur for the life of the asset, and they determine whether it keeps running.

What the evidence says, and where it is contested

ContestedWhether recent industrial policy has been cost effective cannot presently be answered, because the evidence needed does not yet exist. Facilities funded in the early 2020s are still reaching production, obligations are still converting to outlays, and the operating economics that determine survival will not be observable for years. One position holds that construction put in place already demonstrates real activity at scale. A second holds that activity is not the same as capability, and that the test is whether specific things can now be produced that could not be before. Both are arguing from incomplete data, and the disagreement will be settled by plants that either run at rate or do not.

Related Institute research

Reshoring Nitrile Gloves: America’s Cheapest Industrial Vulnerability
What one product actually costs to bring back.

Reshoring, Onshoring, Nearshoring
Why announcement counts and realised capacity are different series.

Why Reindustrialization Requires More Electricity
The lead time that often determines the real schedule.

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 →