Space Manufacturing Measurement and the Hidden Output of the Space Economy

Key Takeaways

• Space manufacturing measurement now needs plant-use data, not just sales data.
• Own-account production may hide satellite and launch vehicle investment.
• Better industrial statistics can improve space policy, finance, and supply-chain planning.

Space Manufacturing Measurement Starts With a 25.2% GDP Share

Manufacturing accounted for 25.2% of U.S. space economy gross domestic product in 2022, making it the largest sector in the U.S. space economy measured in the Bureau of Economic Analysis working paper Measuring Space Manufacturing Plant Utilization and Own-Account Production. The same paper reports that space manufacturing represented 6.4% of U.S. computer and electronic products manufacturing GDP and 7.5% of other transportation equipment manufacturing GDP in 2022. Those figures explain why space manufacturing measurement has become a policy, business, and national-accounting issue rather than a niche statistical exercise.

The larger setting changed again with BEA’s March 2025 release of U.S. space economy statistics for 2012 through 2023. BEA estimated that the U.S. space economy accounted for $142.5 billion, or 0.5% of total U.S. GDP, in 2023. It also estimated $240.9 billion in gross output, $57.9 billion in private-sector compensation, and 373,000 private-sector jobs tied to space economy activity. Manufacturing real gross output grew 2.9% in 2023, even as total real gross output for the space economy declined 0.6%, according to BEA’s March 2025 space economy article.

The BEA paper’s central point is direct: sales and revenue data alone do not show whether space manufacturing plants are operating near capacity, whether new capacity has come online, or whether firms are building capital goods for themselves without a market sale. A satellite factory producing spacecraft for its own broadband constellation may contribute to national investment and future productive capacity even when no arm’s-length sale takes place. A launch vehicle factory can show the same accounting problem when a company builds reusable vehicles for its own operations rather than selling finished rockets to another company.

Space manufacturing includes more than spacecraft assembly. BEA’s space economy definition includes goods and services used in space, goods and services that directly support those used in space, activities that require direct input from space, and activities associated with studying space. That makes computer and electronic products manufacturing part of the space economy when it produces satellites, ground stations, Global Positioning System equipment, and positioning, navigation, and timing equipment. Other transportation equipment manufacturing enters through space vehicles, launch vehicles, and related systems.

The measurement issue has become more pressing because the space economy mixes government procurement, defense and security demand, commercial satellite constellations, launch services, ground equipment, research and development, data services, and wholesale distribution. BEA’s space economy statistics derive from supply and use tables and national income and product accounts, then isolate space-related activity from product categories that often contain both space and nonspace activity. That method works well for a satellite account, but it still depends on the ability to identify where production occurs, how much of it relates to space, and whether market transactions capture the value being created.

Why Plant Utilization Changes the Space Manufacturing Picture

Manufacturing plant utilization measures how much of a plant’s productive capability is being used. A plant operating at 60% utilization has idle capacity. A plant operating near full use may face bottlenecks, labor constraints, supplier delays, or the need for capital expansion. For space manufacturing, this matters because demand can grow in steps. A new satellite constellation, a defense procurement surge, a launch vehicle ramp-up, or a large communications equipment order can raise output before official statistics clearly show whether the industry expanded capacity or strained existing plants.

The BEA paper introduces the Space Economy Manufacturing Plant Utilization Index, or SEMPI. The index combines public data from the Census Bureau’s Quarterly Survey of Plant Capacity Utilization, BEA’s space economy satellite account, Federal Reserve Board data use, and Defense Logistics Agency interests in industrial readiness. The QPC provides statistics on utilization rates for U.S. manufacturing and publishing sectors, and Census describes it as a source for rates of capacity utilization in those sectors.

SEMPI uses North American Industry Classification System codes to connect space-related manufacturing categories with QPC plant utilization data. The paper groups the most relevant manufacturing activity into computer and electronic products and other transportation equipment. Computer and electronic products include communications equipment, electronic components, semiconductors, circuit boards, and related equipment. Other transportation equipment includes aircraft, aircraft engines, guided missiles, space vehicles, propulsion units, and other related parts.

The index does not claim to see every space factory directly. It works through industry categories, available public statistics, and BEA weights for space-related value added. This creates a practical but imperfect tool. For example, the NAICS codes used in public data can combine space and nonspace production. A facility category can include commercial aircraft, missile systems, crew spacecraft, launch vehicle components, or other transportation equipment in the same broad data bucket. That prevents a clean view of one rocket plant or one satellite production line.

The benefit is that SEMPI turns scattered public information into a space-focused utilization measure. It can show whether the space manufacturing base appears stretched, underused, recovering after a shock, or changing its mix between electronics-heavy production and vehicle-heavy production. It can also support better interpretation of price changes. A rise in prices paired with high utilization may point toward capacity strain. A fall in utilization paired with higher real output may point toward new capacity coming online.

The BEA paper reports that space economy manufacturing plant utilization averaged 67.2% between 2012 and 2021. It also reports a persistent decline during 2015, when utilization fell from 76.4% in the first quarter to 65.7% in the fourth quarter. That pattern, combined with evidence of rising real output in separate research by Tina Highfill and Matthew Weinzierl on real growth in space manufacturing output, fits a scenario in which demand encouraged investment in new plant capacity. The paper also identifies a sharp first-quarter 2020 decline, likely tied to COVID-19 effects, followed by recovery later that year.

How SEMPI Connects Space Factories to Supply Chains

SEMPI matters because the space economy has become dependent on manufacturing chains that cross several industrial categories. Satellites require structures, radios, antennas, power systems, computers, sensors, propulsion, software integration, test equipment, and launch adapters. Launch vehicles require engines, tanks, avionics, valves, composite structures, propulsion components, and ground support systems. Ground equipment brings terminals, receivers, antennas, circuit boards, and communications infrastructure into the same industrial story.

A utilization index can help distinguish weak demand from constrained supply. Low utilization may suggest that factories have room to increase output. High utilization may suggest that orders are outrunning available plant capacity. A utilization decline can be bad if it reflects collapsing demand, but it can also occur after firms invest in new production lines, new tooling, or new facilities. In that second case, utilization can fall because the denominator grew faster than output during the transition.

The BEA paper’s interpretation of the 2015 decline points toward that second possibility. Space manufacturing real output rose in the broader period studied by Highfill and Weinzierl, and the paper connects the utilization pattern with rising launch vehicle production or plans. This distinction matters for investors, suppliers, and public agencies because a declining utilization rate does not automatically mean industrial weakness. It can reflect a plant base preparing for higher production levels.

The COVID-era pattern has a different meaning. The BEA paper reports that overall space economy manufacturing utilization fell from 63.8% in 2019Q4 to 54.5% in 2020Q1, with recovery to 64.1% by 2021Q1. That short, sharp decline looks more like a temporary supply shock than a permanent reduction in capacity. The paper reports that computer and electronic products recovered more strongly than other transportation equipment. Electronics manufacturing may have had broader demand channels and greater flexibility than vehicle-heavy production tied to aerospace facilities.

The difference between electronics and transportation equipment also matters for defense and security planning. Space systems often depend on specialized components, secure supply chains, controlled technologies, and mission-specific production processes. A bottleneck in an electronics category can affect satellite buses, payloads, terminals, and ground networks. A bottleneck in propulsion or vehicle manufacturing can affect launch cadence, replenishment capacity, and schedule assurance for civil, commercial, and defense customers.

SEMPI could become more useful if future versions gained finer detail. The BEA paper notes that public data do not always allow distinctions such as avionics versus transponder equipment at useful levels, or guided missiles versus crew spacecraft production at sufficiently narrow industry levels. The measurement problem is not simply a statistics problem. It reflects how industrial classifications group real factories that serve multiple markets.

The following table summarizes the measurement value of SEMPI for space manufacturing users.

Own-Account Production Makes Some Space Output Harder to See

Own-account production occurs when a business produces a capital asset for its own use rather than buying that asset from another company. The BEA paper applies that concept to satellites and space vehicles. If a company builds a launch vehicle and uses it to launch its own payloads, no sale of the vehicle occurs. If a company builds satellites for its own network, no sale of those satellites occurs. Standard revenue and shipment data may miss part of the investment value unless national accountants make an explicit adjustment.

This issue already exists in other parts of the U.S. national accounts. BEA measures own-account software when companies build software for internal use. It also recognizes own-account construction and research and development as investment categories. BEA’s National Income and Product Accounts Handbook describes the concepts and accounting framework behind the national income and product accounts, including the sources and methods used to prepare the estimates.

The paper uses SpaceX as a clear example because it manufactures Falcon 9, a reusable two-stage rocket, and operates the Starlink satellite internet system. SpaceX’s Falcon 9 page describes the vehicle as designed and manufactured by the company for transport of people and payloads into Earth orbit and beyond. The accounting question is not whether an internal asset has economic value. It does. The question is how to estimate that value when no market sale occurs.

If the asset becomes part of the company’s capital stock, the investment side of gross domestic product may be understated without a method for estimating the production cost. That gap may grow as satellite internet systems, reusable launch systems, orbital transfer vehicles, and vertically integrated space firms expand.

The BEA paper distinguishes research and development from production. Prototype development, testing, and materials in the test phase fit within research and development. Once a vehicle or satellite design moves beyond the test phase into production of operational assets, the classification changes. That line can be hard to draw in space because firms often test, revise, and deploy hardware in operational settings. A satellite batch can contain incremental design changes. A reusable vehicle can change between flights. National accountants still need boundaries that can be applied consistently.

Own-account production also matters because the space economy increasingly includes firms that integrate manufacturing, launch, operations, data services, and customer-facing services under one corporate structure. Traditional manufacturing statistics work best when a manufacturer sells a finished product to another firm. Vertically integrated space firms can turn a factory output into an internal network asset, an internal launch capability, or a service platform. The transaction disappears from sales data, but the economic production does not disappear.

Labor, Materials, and Capital Services Need a Cost-Based Method

The BEA paper proposes cost-based measurement as the most practical path for own-account satellite and space vehicle production. The method would estimate labor costs, materials, intermediate inputs, and capital services. That approach mirrors own-account software estimation, which begins with labor categories and then adds other production costs.

Labor is the most visible starting point. The BEA paper uses 2023 Bureau of Labor Statistics Occupational Employment and Wage Statistics for communications equipment manufacturing and aerospace product and parts manufacturing. The paper identifies aerospace engineers, aerospace engineering and operations technologists and technicians, electrical engineers, and assemblers and fabricators as occupations likely to participate in satellite and space vehicle manufacturing. The BLS page for communications equipment manufacturing provides industry-specific occupational employment and wage estimates for NAICS 334200, and the BLS page for aerospace product and parts manufacturing provides comparable industry-specific estimates for NAICS 336400.

The BEA paper gives 2023 occupational detail for two industries. In communications equipment manufacturing, assemblers and fabricators accounted for 11,740 workers, or 14.04% of employment in that industry. Electrical engineers accounted for 2,370 workers. Aerospace engineers accounted for 500 workers. In aerospace product and parts manufacturing, aerospace engineers accounted for 26,400 workers, assemblers and fabricators for 54,400 workers, and electrical engineers for 9,380 workers. Those categories offer a labor-cost foundation, but they do not isolate space work from aircraft, defense systems, terrestrial communications, or nonspace electronics.

Materials create the second measurement problem. Satellites and launch vehicles use specialized inputs, but public data may not identify which purchased materials went into own-account space assets. Analysts can work from representative industry relationships between compensation and intermediate inputs, satellite mass, launch counts, procurement data, or company disclosures. The BEA paper notes that commercial market research has estimated “virtual revenue” for own-account satellites based partly on mass. Public methods would need to avoid proprietary dependence where possible.

Capital services create the third component. A factory uses buildings, machinery, tooling, test stands, clean rooms, software systems, and equipment to produce satellites and vehicles. Own-account production estimates should include the value of those capital services, not just direct wages and purchased materials. That makes the estimate more complex, but it also makes it more aligned with how national accounts treat the cost of producing capital assets for internal use.

The following table compares candidate measurement inputs.

The Measurement Problem Extends Beyond One Company

SpaceX is a useful example, but the measurement problem is broader. Any company that builds satellites, launch vehicles, spacecraft components, ground systems, or orbital assets for its own service business may create output that does not pass through ordinary sales channels. Broadband constellations, Earth observation constellations, in-space logistics services, hosted payload networks, and future commercial space station systems can all create similar accounting questions.

The OECD Handbook on Measuring the Space Economy treats space economy measurement as a task involving definitions, actors, industry surveys, and impact assessment. Its second edition was designed to support public agencies, industry users, and private decision-makers through better data collection and clearer statistical concepts.

International comparison will remain difficult because national statistical systems use different industrial classifications, survey methods, and boundaries for space activity. The BEA method is grounded in U.S. supply and use tables, U.S. national accounts, and U.S. source data. Other countries may focus on agency budgets, direct industry surveys, export classifications, or company-level reporting. These differences matter when policymakers compare national space sectors, industrial capacity, workforce depth, and supply-chain resilience.

A better own-account production method would also change how commercial space growth is interpreted. A vertically integrated company can appear smaller in manufacturing statistics than it really is if its factory output becomes internal capital. A service company can look more service-heavy than its production process suggests. A launch provider can appear to sell services but not the internally built assets that support those services. National accounts need to capture both the service sale and the investment in durable production assets.

Defense and security demand adds another layer. Space manufacturing often supports national security missions, missile warning, communications, navigation, Earth observation, surveillance, and resilience. Some of the industrial base sits in categories that combine civil aerospace, defense production, and space systems. If space-specific capacity cannot be separated from broader aerospace categories, public agencies may struggle to judge whether industry can handle surge demand, replenishment after losses, or parallel civil and defense procurement needs.

The measurement issue also affects inflation analysis. BEA’s March 2025 space economy statistics report that manufacturing prices declined on average from 2012 to 2023, driven by computer and electronic products, before rising in 2023. Price behavior can differ between electronics and vehicle-heavy manufacturing. An electronics segment can benefit from learning curves and semiconductor-driven price changes. A launch vehicle segment may face material, labor, certification, and facility constraints. A single manufacturing headline can hide these different forces.

What Better Space Manufacturing Statistics Would Change

Better space manufacturing statistics would help companies, investors, and public agencies interpret growth more accurately. A manufacturing sector can grow because demand rises, prices rise, productivity improves, quality improves, new capacity appears, or accounting captures hidden production more fully. Each explanation points to a different business or policy response.

For companies, plant utilization data can guide supplier strategy and capital planning. A supplier seeing sustained high utilization in a relevant category may expect longer lead times and higher prices. A prime contractor seeing low utilization may have greater leverage in procurement, but only if the capacity aligns with its technical needs. A satellite operator planning a constellation refresh may need to know whether electronics capacity, propulsion suppliers, or structures manufacturers create the binding constraint.

For investors, better measurement can reduce confusion between value creation and revenue recognition. Own-account production can produce real assets with economic value before those assets generate service revenue. A company building internal satellites may depress near-term margins but raise future capacity. A cost-based measure does not replace company analysis, but it can improve the macroeconomic picture of how much investment the space economy is absorbing.

For public agencies, better statistics support procurement, industrial policy, and emergency readiness. The Defense Logistics Agency uses QPC data to assess whether U.S. industries can meet demand under national emergency scenarios, according to the BEA paper. Space systems now support communications, navigation, weather, missile warning, science, and disaster response. Industrial capacity is part of readiness, and readiness cannot be measured only by annual revenue.

For BEA and other statistical agencies, the paper points toward a broader issue: modern production increasingly crosses the boundaries between manufacturing, software, services, data, and internal capital formation. Space is a clear case because hardware-intensive firms often sell services, not just products. Satellite broadband sells connectivity. Earth observation firms sell imagery, analytics, or subscription access. Launch firms sell transportation services. The physical assets behind those services may come from in-house manufacturing.

As of May 2026, BEA’s space economy page states that the agency will no longer regularly produce these statistics. The same page still hosts the March 2025 release, data files, articles, and prior estimates, but the statement means routine federal publication may not continue under the same schedule. That makes the April 2025 working paper more significant because it identifies measurement methods that could be reused, improved, or adopted by other organizations even if routine federal publication changes.

The next stage of space economy measurement will likely require a mix of public data, confidential microdata, company surveys, regulator filings, launch records, satellite deployment databases, and industry research. Publicly available data can support initial indicators such as SEMPI. More precise estimates may require data access that protects company confidentiality but allows statistical agencies to separate space from nonspace production and market sales from own-account capital formation.

Space Manufacturing Measurement and Policy Decisions

Space manufacturing measurement now sits close to policy decisions about launch infrastructure, spectrum, defense procurement, workforce development, export controls, and supply-chain resilience. A government cannot easily plan industrial policy if the main data show output but not plant pressure. A company cannot interpret market demand well if a competitor’s internally produced satellites create capacity that never appears as product sales. A workforce planner cannot judge the skill base if occupation data do not separate space-specific work from broader aerospace and electronics employment.

The BEA paper’s plant utilization work points toward one practical answer: build indicators from existing public data, explain their limits, and improve granularity over time. The own-account production section points toward a second answer: use cost-based methods where market transactions do not exist. Together, those approaches move space economy measurement from simple revenue tracking toward a fuller view of production, capacity, and investment.

The approach also gives analysts a better way to think about commercial constellations. A broadband satellite network is partly a telecommunications service, partly a manufacturing program, partly a launch-demand driver, partly a ground-equipment market, and partly a capital investment project. The economic footprint appears in several places at once. Without a satellite account, those links are scattered. Without utilization and own-account production measures, parts of the same system remain undercounted or hard to interpret.

A practical measurement agenda would start with the categories already identified in the paper: computer and electronic products and other transportation equipment. It would then refine the space share within each category, separate satellite and ground equipment where possible, and explore ways to distinguish space vehicles from aircraft and missile systems in public-facing statistics. Confidential data could help build better public aggregates without exposing company-level details.

The same agenda would improve workforce estimates. BEA’s September 2025 working paper The Space Economy Workforce and STEM Occupations states that BEA estimated 373,000 private-sector space economy employees in 2023 and that those employees spanned almost all industry sectors of the U.S. economy. It also reports that only NAICS 3342, communications equipment manufacturing, had more than 40% of total gross output that was space related. That finding supports the April 2025 paper’s warning that many industries contain both space and nonspace activity.

A better statistical system would not remove uncertainty. Space firms will keep changing business models, and industrial classifications will always lag new production patterns. The value of SEMPI and own-account production methods is that they give analysts a structured way to ask better questions: whether factories are constrained, whether capacity is expanding, whether internal assets are missing from GDP, and whether the visible market understates the industrial base behind space services.

Summary

Space manufacturing is large enough inside the U.S. space economy to deserve more detailed measurement than ordinary sales data can provide. The BEA working paper shows why plant utilization and own-account production matter. SEMPI gives analysts a way to connect public plant capacity data with BEA’s space economy categories. Own-account production analysis explains why internally built satellites and space vehicles can create real investment even when no sale occurs.

The main policy lesson is that space economy growth cannot be read from revenue alone. A space company may build internal capital assets, use plants at changing levels, expand capacity before sales arrive, or shift the mix between electronics and vehicle-heavy production. Public statistics need methods that can detect those patterns without relying on one company’s disclosures or one market forecast.

Better measurement would support better decisions in procurement, finance, workforce planning, supply-chain management, and national security. It would also help separate industrial expansion from price changes, temporary shocks, and classification artifacts. The BEA paper does not solve every measurement problem. It gives the space economy a more useful statistical direction: measure capacity, measure internal production, and treat space manufacturing as a full industrial system rather than a set of isolated sales.

Appendix: Useful Books Available on Amazon

• The Space Economy
• The Space Barons
• Rocket Billionaires
• Space Is Open for Business
• Escaping Gravity
• OECD Handbook on Measuring the Space Economy

Appendix: Top Questions Answered in This Article

What Is Space Manufacturing Measurement?

Space manufacturing measurement is the process of estimating the economic activity tied to producing satellites, spacecraft, launch vehicles, ground equipment, navigation equipment, and related systems. It includes normal sales-based statistics, but it also needs measures of plant capacity, production limits, price changes, internal capital formation, and industry mix.

What Is SEMPI?

SEMPI is the Space Economy Manufacturing Plant Utilization Index introduced in the BEA working paper. It combines public plant utilization statistics with BEA space economy weights to estimate how much of the relevant U.S. manufacturing plant base is being used for space-related production.

Why Does Plant Utilization Matter?

Plant utilization helps identify whether production is constrained by factory capacity or whether plants have room to increase output. A high utilization rate can point toward bottlenecks or the need for new investment. A lower rate can mean weak demand, but it can also mean new capacity has been added.

What Is Own-Account Production?

Own-account production occurs when a company builds a capital asset for its own use rather than buying it from another company. In the space economy, this can include satellites, launch vehicles, or related systems built internally for use in the company’s own network or service business.

Why Can Own-Account Production Lead to Undercounting?

Standard economic statistics often rely on sales, shipments, or revenues. When a company builds an asset for itself, no market sale occurs. If national accounts do not estimate the value of that internal production, investment and GDP can be understated.

Which Industries Matter Most for Space Manufacturing?

The BEA paper focuses on computer and electronic products and other transportation equipment. The first category includes communications equipment, electronic components, semiconductors, circuit boards, satellites, and ground equipment. The second category includes space vehicles, launch vehicles, propulsion systems, and related aerospace equipment.

Why Are NAICS Codes a Limitation?

NAICS codes group economic activity by industry, but public categories can be broader than the space activity being studied. A single category may include aircraft, missiles, spacecraft, electronics, and nonspace products. This limits how precisely analysts can isolate space-specific manufacturing capacity.

How Did COVID-19 Affect Space Manufacturing Utilization?

The BEA paper reports a sharp decline in space economy manufacturing plant utilization in 2020Q1, followed by recovery later that year. Other transportation equipment experienced a deeper shock than computer and electronic products, suggesting that different parts of the manufacturing base responded differently.

Why Does This Matter for Defense and Security?

Defense and security space systems depend on industrial capacity, specialized suppliers, electronics, propulsion, launch vehicles, and replenishment ability. If public data cannot show where capacity is strained, procurement planning and emergency readiness become harder to assess.

What Could Improve Space Manufacturing Statistics?

Better statistics would use more detailed industry data, clearer space shares inside mixed industries, cost-based methods for own-account production, confidential data protections, and better separation between research and development, prototype work, and operational production.

Appendix: Glossary of Key Terms

Space Manufacturing Measurement

Space manufacturing measurement refers to the statistical process of estimating production, capacity, prices, employment, investment, and industrial activity tied to space-related goods. It covers direct space products such as satellites and launch vehicles, plus supporting equipment such as ground systems and navigation hardware.

Gross Domestic Product

Gross domestic product is the value of final goods and services produced within an economy during a set period. In a space economy satellite account, GDP measures the value added by industries producing goods and services connected to space activity.

Gross Output

Gross output measures the total value of goods and services produced by an industry, including intermediate products later used by other industries. It is broader than GDP because it includes the value of production before subtracting intermediate inputs.

Space Economy Satellite Account

A space economy satellite account reorganizes existing national economic data to show space-related production and spending that would otherwise be spread across many industries. It does not replace the main national accounts; it adds a focused space-economy view.

Plant Utilization

Plant utilization measures the share of a manufacturing facility’s productive capability that is being used. A utilization rate can help analysts identify bottlenecks, idle capacity, demand shifts, temporary disruptions, or new investment in factory capability.

SEMPI

SEMPI stands for Space Economy Manufacturing Plant Utilization Index. It is a BEA working-paper method that combines public plant utilization data with space economy weights to estimate manufacturing utilization for space-related industrial activity.

Own-Account Production

Own-account production occurs when a business produces a capital asset for its own use. In the space economy, this can include internally built satellites or reusable launch vehicles that support a company’s own service business.

Capital Services

Capital services are the productive services provided by assets such as buildings, machinery, tools, test equipment, and software systems. Cost-based estimates of own-account production need capital services because those assets contribute to production.

NAICS

The North American Industry Classification System is a statistical classification system used to organize businesses by industry. It is useful for economic measurement, but broad categories can combine space and nonspace activity in the same code.

Computer and Electronic Products

Computer and electronic products is a manufacturing category that includes communications equipment, semiconductors, electronic components, circuit boards, and related systems. In the space economy, it includes products such as satellites, ground stations, and navigation equipment.

Other Transportation Equipment

Other transportation equipment is a manufacturing category that includes aircraft, space vehicles, launch vehicles, propulsion units, and related parts. Public data can combine space and nonspace products within this broader category.

Occupational Employment and Wage Statistics

Occupational Employment and Wage Statistics is a Bureau of Labor Statistics program that estimates employment and wages by occupation and industry. It can help estimate labor costs in industries connected to own-account satellite and launch vehicle production.