# Manufacturing investment within the US national accounts

Economy Study · United States

Factory construction is evidence of an investment process, but its economic meaning depends on the question being asked. Spending on a manufacturing site can support construction activity before the factory produces a saleable product. Production machinery can be delivered in a different period from the building work. Software and research can support the same production process without appearing in a construction total. An assessment that moves directly from a building expenditure figure to manufacturing capacity consequently omits several necessary relationships.

The U.S. statistical system provides measures for these different stages. They should be connected, with their boundaries retained, rather than treated as interchangeable estimates of a single manufacturing investment total. This study establishes that connection through the official definitions and a transparent project illustration. It does not estimate the current size of a construction boom or the effect of an investment incentive.

## Three different statistical objects

The Census Bureau’s [construction definitions](https://www.census.gov/econ_file/construction/c30/definitions.html) distinguish work installed or erected at a site from the total financial commitment to a project. The construction measure includes specified building services and site work but excludes land acquisition and most production machinery. Some equipment integral to a structure receives a different treatment from machinery used to make the eventual product. The manufacturing construction category concerns manufacturing sites; it is not an inventory of every asset owned by a manufacturing company.

BEA’s [NIPA Handbook, chapter 6](https://www.bea.gov/resources/methodologies/nipa-handbook/pdf/chapter-06.pdf) addresses private fixed investment across structures, equipment and intellectual property products. Its principal NIPA presentation classifies the investment by product type. The chapter identifies separate fixed-asset estimates by industry. A structure category and an industry classification are therefore different ways of organizing an investment record.

The Federal Reserve’s [G.17 explanatory material](https://www.federalreserve.gov/Releases/G17/About.htm) concerns industrial production and capacity. Production measures real output; capacity addresses sustainable potential output; utilization relates production to capacity. None is simply the dollar value of recent building work. These definitions supply three statistical objects: construction performed, assets acquired or produced, and productive activity or potential. The analytical task is to explain their connection in time.

## A project can generate expenditure before capacity

Consider a hypothetical manufacturer building a new facility. In the first year, contractors prepare the site and erect part of the structure. The manufacturer also orders production machinery. In the second year, building work continues and equipment arrives. Testing, customer qualification and operational preparation follow. The project can create recorded investment and demand for construction services while the plant’s commercial output remains zero.

This sequence need not indicate failure. Expenditure and output belong to different stages of the investment. Nor does a completed building prove that all remaining stages will succeed. A missing power connection, a delayed machine or a customer qualification problem can postpone production after substantial capital has been committed. The relevant assessment is whether the project is progressing through the stages needed for its intended use.

An investment monitoring system should therefore associate the expenditure with a project stage. It should also retain the difference between the contractor performing the work, the enterprise owning the asset and the industry eventually using it. Those relationships can coincide, but they need not do so. A description of the expenditure purchaser cannot automatically establish the industry in which the resulting services will be used.

This matters particularly when several factories are at different stages. Aggregate construction can decline as a large group of plants moves toward equipment installation and operation. The decline can coexist with an increase in new production. Conversely, accelerated building work can precede any comparable increase in output. The sign of a construction change is not, by itself, the sign of a capacity change.

## Reconcile the expenditure before interpreting its growth

The following values are assumed and describe no actual U.S. project or data release. They illustrate how the composition of expenditure changes an investment assessment. All amounts use the same arbitrary monetary unit.

| Project expenditure | Period A | Period B | Assumed price change |
| --- | ---: | ---: | --- |
| Structure work | 120 | 180 | Prices rise 25 percent |
| Production equipment delivered | 80 | 70 | Prices unchanged |
| Software qualifying as investment | 10 | 15 | Prices unchanged |
| Total of the three stated categories | 210 | 265 | Component prices differ |

Nominal structure expenditure rises by 50 percent. Total nominal expenditure across the three categories rises by 55, or approximately 26.19 percent. Neither number describes the change in the volume of the entire project’s investment.

At period-A prices, the period-B structure work is 180 divided by 1.25, or 144. Under the example’s unchanged equipment and software prices, the three period-B amounts at those same fixed prices sum to 229. Relative to 210, that is an increase of approximately 9.05 percent. The volume of structure work alone rises by 20 percent.

These calculations deliberately use a single fixed-price reference for a small illustrative account. They do not construct BEA’s published Fisher quantity indexes, and they do not justify adding published chained-dollar components. The [separate data note on chained dollars](/research/chained-dollar-components/) explains that limitation.

The example establishes a narrow result. A large nominal increase in one component can overstate both its own volume increase and the volume increase of a wider investment bundle. It does not establish that actual U.S. construction prices rose by the assumed amount, or that equipment spending offset building growth in any particular period. Those are empirical questions requiring the relevant price and expenditure series.

A further assumed land purchase of 50 should not be added as though it were newly produced structure work. Similarly, a loan of 100 to finance the project should not be added to the asset expenditure. Doing so would count the financing transaction alongside the use of funds and call both productive investment. The economic assessment needs a financing record, but it should remain a separate record.

## Imports and industry ownership change the interpretation

Suppose 40 of the period-B equipment expenditure purchases imported machinery. The machine can become an asset used in U.S. production even though its manufacture occurred abroad. That distinction prevents two opposite errors: excluding an imported asset from the domestic capital assessment, or treating its full purchase price as newly produced U.S. output.

BEA’s chapter 6 explicitly includes domestic purchases of imported equipment in private fixed investment. In expenditure accounting, the imported content must also be recognized through the external account. The implication for analysis is to retain both the investment use and the origin of production. A domestic factory expansion can draw on foreign production capacity during its construction and equipment phase.

Ownership creates another boundary. BEA’s [fixed-asset definitions](https://www.bea.gov/resources/learning-center/definitions-and-introduction-fixed-assets) explain the ownership basis used in distributing assets across industries, including the treatment of operating and capital leases. An asset’s statistical owner and its physical user can therefore require separate attention. A researcher assembling an industry estimate should read the classification rule before allocating every leased asset to the establishment where it is used.

The same caution applies to a building sold after completion. An asset transfer and the creation of new productive assets are different events. A transaction can change ownership or financing without recreating the structure. The relevant account must distinguish new formation from transfers and from the services associated with those transfers.

## Capacity requires its own evidence

The project illustration contains no capacity estimate. Converting its expenditure into potential output would require information about the technology, product mix, equipment, operating schedule and complementary inputs. A spending-to-capacity conversion borrowed from another industry would introduce assumptions that the expenditure data cannot test.

Capacity utilization also needs a denominator. In a second hypothetical illustration, output rises from 80 to 84 while capacity rises from 100 to 110. Utilization falls from 80 percent to approximately 76.36 percent even though output increases by 5 percent. The fall is compatible with the expansion of available capacity; it does not establish a contraction in production.

For an individual plant, a utilization measure might additionally require a distinction between nameplate capacity, qualified capacity and output sustainable under actual operating conditions. Those concepts should not be merged merely because each can be expressed in physical units. For the aggregate U.S. series, the Federal Reserve’s definition and estimation methods govern the published measure.

A useful account consequently places expenditure, completion, capacity and production on a common timeline without requiring them to move together. Differences among them can reveal where the investment process stands. They become a statistical problem only if the measures are treated as equivalent or combined outside their definitions.

## Implications for an industrial investment assessment

The first finding is that manufacturing construction is a component and a stage of the investment process. A complete capital assessment needs equipment and intellectual property as well as structures, within a consistent ownership and geographic boundary. Adding financial commitments or land transactions to a construction number does not repair missing asset coverage.

The second is that nominal expenditure growth needs a price and composition assessment. A structure-heavy increase can reflect a different volume change from an equipment-heavy increase of the same nominal size. The appropriate real aggregate must be constructed or obtained under a coherent method.

The third is that operating results require independent evidence. A completed investment can enlarge potential production, replace an old facility, change product quality or reduce costs. Which result occurred cannot be established from construction expenditure alone. Project records and production measures should be used to answer that question.

These distinctions also limit policy attribution. The correct measurement of investment is necessary to evaluate support, but it does not show what would have happened without that support. The [report on additionality](/research/industrial-investment-additionality/) addresses that separate counterfactual question. Keeping measurement and attribution distinct allows each part of an industrial investment assessment to carry the evidence it actually supports.

---

American Economic Senate