Last Updated: September 9, 2026
Manufacturing Employment by Industry: 2026 Jobs and Pay Report
Manufacturing is often discussed as one block of 12 million-plus workers. That hides the real labor market. Transportation equipment and food manufacturing each employ more than 1.7 million people, while leather manufacturing employs about 25,000. Pay ranges from roughly $54,000 in textile product mills to more than $153,000 in computer and electronic products.
This report ranks all 21 three-digit manufacturing subsectors using the BLS Quarterly Census of Employment and Wages 2024 annual national file. Every job, establishment, payroll, pay, growth, and calculated plant-density figure links to its original source.
private manufacturing jobs, BLS QCEW
Manufacturing Employment by Industry Table
The table ranks industries by annual average employment. BLS QCEW covers more than 95% of US jobs through unemployment insurance records, according to the QCEW program handbook. The figures below are private-sector annual averages, which avoids mixing a single survey month with a full-year payroll measure.
The 21 rows contain 168 directly reported or calculated data points: jobs, job share, establishments, workers per establishment, annual pay, payroll, and employment change. The source file also includes weekly wage and quarterly fields for analysts who need a deeper time series. Download the BLS QCEW files to reproduce or extend the table.
The Industries That Employ the Most Manufacturing Workers
Transportation equipment and food are nearly tied
Transportation equipment employed 1,770,605 people, only 3,940 more than food manufacturing. The categories have very different footprints. Transportation equipment averaged 94.6 workers per establishment, while food averaged 42.1. Automotive assembly, aerospace, shipbuilding, rail equipment, and parts production tend to concentrate headcount in larger facilities. Food production stretches across meat processing, dairies, bakeries, grain mills, seafood, produce, and specialty foods.
The near tie matters to suppliers. A seller looking only at job totals might treat the markets as similar. Account strategy says otherwise. Food has more than twice as many establishments, so it offers more individual buying locations. Transportation equipment has much higher headcount per site, which can support larger contracts for automation, staffing, safety, maintenance, software, and plant services.
Fabricated metals has the widest account universe
Fabricated metal products employed 1,437,729 workers at 60,906 establishments. No other subsector has as many locations. The average works out to 23.6 jobs per establishment, showing why this market feels fragmented. It includes machine shops, sheet metal fabricators, structural metal producers, coating operations, spring makers, fastener plants, and other specialists.
That structure changes selling economics. A national campaign can reach a large number of plants, but average deal sizes may be smaller than in aerospace or automotive. Good segmentation needs at least NAICS detail, employee count, process, geography, and certification. A generic list of “metal companies” mixes five-person job shops with multi-site suppliers and creates a weak sales territory.
Machinery and electronics round out the top five
Machinery employed 1,106,882 people, while computer and electronic products employed 1,019,785. Machinery spans farm, construction, industrial, engine, pump, compressor, tool, and material-handling equipment. Electronics covers semiconductors, instruments, communications equipment, computers, and electronic components.
Electronics stands apart on compensation. Its average annual pay was $153,352, the highest in the table. That reflects its mix of engineering, technical, research, software-adjacent, and precision production work. The sector also recorded a 6.7% employment decline in the annual QCEW comparison, a reminder that high pay and job growth do not always move together.
Manufacturing Pay by Subsector
Pay differs by nearly three to one across manufacturing industries. Computer and electronic products led at $153,352 in average annual pay. Petroleum and coal products followed at $141,329, and chemicals paid $118,180. These are industry averages across all covered workers, not starting wages for production operators.
Capital intensity, occupational mix, location, shift patterns, union coverage, hazard exposure, and technical requirements all influence the number. A refinery employs process engineers, instrument technicians, operators, maintenance trades, laboratory staff, and safety professionals around equipment that runs continuously. Semiconductor and instrument manufacturers employ a higher share of engineers and specialized technicians. Those mixes lift industry averages.
At the other end, textile product mills averaged $54,390, apparel averaged $57,949, furniture averaged $59,317, and leather products averaged $59,154. These numbers are useful for broad market comparisons, but employers should price an actual opening with occupation and local-area data from the BLS Occupational Employment and Wage Statistics program.
Annual payroll reveals a different kind of market scale. Electronics paid $156.4 billion, transportation equipment paid $165.1 billion, food paid $108.5 billion, fabricated metals paid $105.3 billion, chemicals paid $105.6 billion, and machinery paid $97.2 billion. Payroll indicates workforce purchasing power and the scale of employer spending, but it is not total labor cost because benefits are excluded.
Employment Growth and Contraction
Only four subsectors in the table posted positive 2024 annual employment change. Petroleum and coal products led at 2.5%, food increased 1.9%, electrical equipment and appliances increased 1.6%, and transportation equipment edged up 0.1%. Growth in electrical equipment fits the continuing buildout of batteries, switchgear, motors, transformers, and grid equipment, though one annual observation should not be treated as a permanent forecast.
Apparel recorded the steepest decline at 6.9%, followed by electronics at 6.7% and textile mills at 5.8%. Printing, leather products, furniture, and textile product mills also contracted by more than 3%. Long-running trade exposure, automation, demand shifts, consolidation, and plant location decisions can all contribute, but QCEW itself reports employment, not the cause of change.
Do not read a negative national rate as proof that every company is shrinking. A subsector can lose employment nationally while individual plants add shifts, open lines, or struggle to fill technical roles. The BLS manufacturing industry page provides current monthly employment context, while QCEW supplies the more detailed annual benchmark. Use both when timing a hiring, expansion, or sales decision.
For workforce planning, the direction of travel is only one input. Retirement exposure, local training capacity, occupation mix, turnover, overtime, and commuting patterns determine whether a plant can hire. A flat industry with an aging skilled-trades workforce can produce more urgent recruiting demand than a fast-growing industry with a deep labor pool.
Jobs per Manufacturing Establishment
Workers per establishment is a simple ratio, not a BLS-published average plant size. It divides annual average employment by annual average establishment count from the same QCEW source file. The measure gives sales teams and economic developers a quick look at how concentrated employment is across locations.
Transportation equipment had the highest ratio at 94.6 jobs per establishment. Paper followed at 62.4, primary metals at 61.2, and petroleum and coal products at 42.6. These industries depend on capital-heavy equipment, continuous processes, large production assets, or integrated plants. Their median facility may differ from the calculated mean because a small number of very large sites can pull the average upward.
Printing averaged 12.8 workers per establishment, textile products 15.5, miscellaneous manufacturing 16.4, and leather products 16.3. The lower ratios point to more fragmented location markets. They can favor local service coverage, distributor networks, lower-ticket offers, and marketing programs that can qualify many smaller accounts efficiently.
Establishment count is not company count. The BLS methodology generally treats a physical location as an establishment, so one manufacturer operating eight plants can appear eight times. That is often helpful for territory design because purchasing, maintenance, safety, and staffing decisions may happen at the plant level even when corporate procurement sets the vendor list.
How to Use This Data
For manufacturers and workforce teams
Benchmark your industry before benchmarking the whole factory economy. A food plant competing for maintenance technicians has a different wage ceiling and shift environment than an electronics facility. Start with the industry average, then move to occupation and metro data. The table can also flag subsectors where high payroll and large plants make retention, training, and automation decisions especially expensive.
For suppliers and industrial marketers
Use jobs as a scale signal and establishments as an account-count signal. High employment plus a low establishment count suggests fewer, larger targets. A high establishment count with fewer workers per plant suggests a broad small and midsize market. Pair the ratios with the Census NAICS definitions so your campaign does not mix industries with different processes and buying problems.
The language on a capability page should match the subsector. A food processor cares about washdown, sanitation, traceability, changeovers, and uptime. A medical-device manufacturer inside miscellaneous manufacturing cares about validation, documentation, clean production, and supplier controls. Transportation equipment suppliers care about program timing, PPAP, traceability, and delivery performance. Employment data identifies the market, but plant-level research tells you what to say.
For economic development and location work
National industry data is the baseline. State and county QCEW files show local specialization and can reveal whether a region has the supplier base, occupations, and adjacent plants an expansion needs. The Census Annual Survey of Manufactures adds shipments, materials, capital expenditures, and other operating measures. BEA GDP by industry adds value added and output.
Employment should not be used alone to rank industry importance. Petroleum manufacturing has relatively few jobs but huge throughput. Fabricated metals has a large workforce and many small plants. Electronics has high pay and research intensity. Food has broad geographic coverage because production often sits near farms, consumers, logistics routes, or cold-chain infrastructure. Each metric answers a different question.
National Manufacturing Context
The National Association of Manufacturers reports roughly 12.6 million US manufacturing workers, more than 239,000 manufacturers, $2.96 trillion in manufacturing value added in the fourth quarter of 2025, and $106,691 in average annual manufacturing compensation including benefits. The QCEW total on this page differs slightly because it covers private employment using a specific 2024 annual file and sums three-digit industries.
The NIST Annual Report on the US Manufacturing Economy connects employment with productivity, output, trade, compensation, and research. That wider view matters. A sector can produce more with fewer workers when equipment, software, process improvement, and product mix raise output per hour. Employment loss therefore does not translate mechanically into an equal decline in real output.
The dataset is most useful as a map of workforce structure. It shows which industries are labor-heavy, which operate through larger plants, which support the biggest payrolls, and where annual job counts were moving. It does not measure contractors, self-employed workers, upstream logistics, construction, utilities, or induced jobs supported by factory spending.
How This Industry Report Fits Other Employment Statistics
The Bureau of Labor Statistics publishes several products that answer different questions. QCEW is the source for the detailed industry employment data here. Current Employment Statistics, often called the establishment survey, supplies faster seasonally adjusted estimates of nonfarm payrolls, average hourly earnings, and production and nonsupervisory hours. The Current Population Survey measures the labor force, unemployment rate, and participation rate through a household survey. These data series should be used together, not substituted without checking definitions.
The monthly Employment Situation from BLS is the primary source for current total employment and the unemployment rate. News coverage may say the unemployment rate ticked higher or that manufacturing rose in one month, but a single monthly estimate is not the same as annual employment in industry. Seasonally adjusted CES numbers are designed for monthly change. QCEW annual averages are better for comparing each major industry sector.
Long-term historical data add perspective. The St. Louis Fed FRED manufacturing employment series shows about 19.5 million in 1979 and a pandemic-era low near 11.4 million in April 2020. Manufacturing employment has declined from that late-1970s peak even as real output and labor productivity increased over long periods. Output per worker can rise when a plant produces more goods and services with fewer labor hours per unit of output.
Comparisons across 2021, 2022, 2023, 2024, 2025, and 2026 also require care. The annual table is anchored to 2024. The 2023 figure is the comparison year for job change. Later 2025 national context comes from NAM and BEA, while 2026 is the report edition. Mixing those dates into one unlabeled number would create a false trend.
Average hourly earnings are different from annual pay and total compensation. Hourly earnings from CES focus on payroll earnings and can be split for production and nonsupervisory workers. Annual pay in QCEW is total wages divided by average employment. Total compensation can include employer-paid benefits. Labor cost decisions need the appropriate measure for the job, industry, location, and benefit package.
The manufacturing sector includes both durable goods and nondurable goods. Fabricated metal products, machinery, electronics, and transportation equipment are durable categories. Food, chemicals, paper, petroleum, textiles, apparel, and leather include major nondurable categories. NAICS is the North American industrial classification used to group these activities. Chemical manufacturing includes businesses where chemical transformation is a core production process.
State employment and the share of total employment vary sharply with local industry concentration. National industry employment cannot show whether a specific metro has toolmakers, controls engineers, or process operators available. Researchers should combine QCEW geography files with OEWS occupation estimates, Census Bureau business data, manufacturing job openings, and local workforce information.
Broader industrial development depends on more than the percentage of workers in factories. Investment in manufacturing, foreign direct investment, research and development, supply chains, energy, infrastructure, and training all shape capacity. The World Bank manufacturing value-added indicator helps with international manufacturing comparisons, while BEA reports US value added and real output in measures such as chained 2017 dollars.
Per capita measures can help compare regions of different sizes, but they should not replace absolute account counts for sales planning. Manufacturing remains a large part of the U.S. economy because its output, supplier purchases, payroll, exports, research, and capital spending extend beyond direct manufacturing jobs. The direct-job table is a benchmark for the manufacturing workforce, not an estimate of every supported job.
Methodology and Limitations
We filtered the 2024 annual BLS QCEW national file to private ownership and the 21 three-digit NAICS manufacturing subsectors. Employment is annual average employment. Establishments are annual average establishment counts. Payroll is total annual wages. Annual pay and year-over-year employment change are QCEW fields. Workers per establishment and employment share are calculations from those source fields.
QCEW suppresses some detailed cells to protect confidentiality, but national three-digit totals are available. Revisions, NAICS changes, late employer reports, and differences between survey programs can produce small discrepancies. CES provides faster monthly estimates. QCEW provides slower but much broader administrative coverage. Census programs use their own business universes and reporting definitions.
The 189 reported and calculated data points on this page should be cited as 2024 annual data, even though the report was published in 2026. Do not relabel the underlying observations as 2026 employment. “2026” identifies the edition and update date, while each paragraph and table explains the observation year.
Frequently Asked Questions
How many people work in US manufacturing?
The 21 private manufacturing subsectors in the 2024 annual BLS QCEW file sum to 12,733,349 annual average jobs. Other published totals may be slightly different because CES is a monthly survey, QCEW is an administrative census, and some national figures include different ownership categories or reference periods. Always name the program and date when quoting an employment total.
Which manufacturing industry employs the most people?
Transportation equipment ranked first in this dataset with 1,770,605 workers. Food manufacturing was close behind with 1,766,665 workers. Fabricated metal products ranked third with 1,437,729 workers. Together, these industries employ people across assembly, machining, fabrication, quality, maintenance, logistics, supervision, engineering, and support functions.
Which manufacturing industry pays the most?
Computer and electronic product manufacturing had the highest average annual pay at $153,352. Petroleum and coal products ranked second at $141,329, followed by chemicals at $118,180. These averages include production and nonproduction employees, so they should not be confused with the wage for one occupation.
What is the difference between an industry and an occupation?
An industry describes what the employer makes. An occupation describes what the worker does. A welder can work for a fabricated metal company, machinery builder, transportation equipment supplier, or furniture manufacturer. An accountant can also work in any of those industries. Use QCEW for industry employment and the BLS OEWS program for employment and wages by occupation.
Does manufacturing employment include contract workers?
Workers are generally counted in the industry of the employer reporting unemployment insurance wages. A temporary worker assigned to a factory may therefore be counted under employment services rather than manufacturing. Independent contractors and self-employed owners are also outside most QCEW covered employment. This is one reason the number of people physically working in plants can differ from published manufacturing payroll employment.
Why use annual average employment?
Annual averages smooth seasonal peaks, shutdowns, strikes, model changeovers, holiday production, and month-to-month survey noise. They are well suited to comparing subsector scale. For current turning points, pair the annual benchmark with the BLS Current Employment Statistics series, which is published monthly and revised as new information arrives.
Turn Industry Data Into Qualified Manufacturing Leads
Learn the system we use to find the right industrial accounts, build useful content, and create a steady RFQ pipeline.
Get the Free Qualified Leads Course →Cite This Resource
Suggested citation
Manufacturing Lead Generation. “Manufacturing Employment by Industry 2026: Jobs and Pay Report.” Updated September 9, 2026. https://manufacturingleadgeneration.com/manufacturing-employment-by-industry/
