The manufacturing skills gap has a small set of numbers that get quoted everywhere. 2.1 million unfilled jobs. A trillion dollars. An aging workforce on the edge of retirement. They show up in vendor decks, conference keynotes, and roughly every LinkedIn post about hiring techs.
Most of them trace back to real studies. The trouble is that the numbers get pulled out of context, mixed across study years, and reframed as facts about maintenance specifically when they describe all of manufacturing.
This page sorts the defensible figures from the folklore. Each row gets a Reliable Confidence Score based on how well it traces to a named primary source, whether that source is current, and whether the number actually measures what people claim it measures.
What these numbers measure, and where they come from
Three bodies of work supply almost every skills-gap statistic in circulation.
The first is the Deloitte and Manufacturing Institute talent study series, which has run roughly every few years for two decades. The Manufacturing Institute is the workforce and education affiliate of the National Association of Manufacturers (NAM). These studies combine a survey of manufacturing executives with an economic model that projects future labor supply and demand. The headline shortfall figures (2.4 million, 2.1 million, 1.9 million) come from this model. They’re projections, not counts of jobs sitting open today.
The second is NAM’s quarterly Manufacturers’ Outlook Survey. This measures sentiment: what share of manufacturers rank a given issue as a top concern. It tells you how worried the industry is, not how many positions are vacant. Workforce led that list for years. By 2026 it had slipped behind trade uncertainty, healthcare costs, and input costs, which matters when you cite it.
The third is the U.S. Bureau of Labor Statistics, which publishes hard employment counts, ten-year occupation projections, and wage data. This is the cleanest source for anything maintenance-specific, because BLS tracks industrial machinery mechanics, machinery maintenance workers, and millwrights as a defined occupation group.
Keep those three straight and most of the confusion clears up. A projection, a sentiment survey, and an employment count are three different things, and they get quoted as if they’re interchangeable.
| Source or Claim | Figure | Reliable Confidence | What It Really Means |
|---|---|---|---|
| 2024 Deloitte and Manufacturing Institute Talent Study | Net need for as many as 3.8M workers (positions to fill) 2024–2033; about half (1.9M) could go unfilled | Highcurrent, primary, methodology disclosed | The most recent figure in the series and the one to cite as “today’s” number. The report’s Figure 2 breaks the 3.8M into 2.8M from retirements, 760,000 from industry growth, and 230,000 from federal investment (IIJA, IRA, CHIPS), so roughly three-quarters is replacing retirees, not new growth. The 1.9M unfilled is a modeled scenario, not a vacancy count. |
| 2021 Deloitte and Manufacturing Institute study | 2.1M unfilled by 2030; up to $1T cost in 2030 | Mediumreal but superseded | The most-quoted number on the internet, and now out of date. The 2024 study replaced it. It was a legitimate projection for its 2021 vintage; using it as today’s figure is the dating error, not the number itself. The $1T is a single-year 2030 economic projection. |
| 2018 Deloitte and Manufacturing Institute study | 4.6M jobs to fill 2018–2028 (1.96M growth + 2.69M retirements); 2.4M could go unfilled; ~$454B at risk in 2028 | Mediumhistorical, superseded by later studies | Useful for trend context, not for current claims. Its forecast window runs through 2028 and a newer study has replaced it, so treat it as background. The “4 million jobs” half-memory people cite usually comes from here (the 4.6M to fill). |
| NAM Manufacturers’ Outlook Survey | Attracting and retaining talent a top concern in most quarters from Q4 2017 through 2024 (65%+ in Q1 2024); not in the top three by 2026 | Highrecurring primary survey | Measures executive sentiment, not vacancies. Strong evidence the concern was real and durable for years. As of Q1 and Q2 2026, trade uncertainty, healthcare costs, and input costs rank above workforce on NAM’s list, so the 65% figure is historical, not the current ranking. |
| BLS, Industrial Machinery Mechanics, Machinery Maintenance Workers, and Millwrights | 538,300 jobs (2024); projected +13% 2024–2034; ~54,200 openings per year; median wage $63,510 (May 2024) | Highgovernment data, directly verifiable | The cleanest maintenance-specific anchor. Covers all industries, not manufacturing alone, and all three roles in the group, including millwrights. BLS says many of the annual openings come from replacing workers who transfer to other occupations or leave the labor force, including retirement. Openings are projected hiring demand, not a count of unfilled jobs. |
| Median age of the manufacturing workforce | About 44 (recent CPS readings near 43.9) vs 42.1 for all workers (2025 CPS); roughly a quarter are 55 or older | HighBLS CPS, 2025 | The workforce is older than average, but only by about two years, and the gap has generally remained modest. The exact decimal drifts year to year. The “44.7 and widening” version floating around traces to 2012 data quoted as current. |
| A maintenance-specific skills gap number | No authoritative national estimate found in the sources reviewed | Highnarrow negative finding | We found no study that isolates a “maintenance skills gap” as its own quantity. Every headline shortfall number is whole-sector manufacturing. Maintenance roles are a subset, sized through BLS occupation data, not a standalone study. |
| “$47M a year lost to poor knowledge sharing” (Panopto/YouGov, 2018) | Modeled cost for an average ~17,700-employee firm | Lowvendor-generated, cross-industry | From the Panopto Workplace Knowledge and Productivity Report, a YouGov survey of about 1,000 U.S. knowledge workers across all industries. It’s a modeled per-large-firm estimate, not a manufacturing or maintenance finding. Useful as illustration of knowledge-loss risk, never as a benchmark. |
The Big Takeaway
The skills gap is real, but the famous number is outdated.
The figure almost everyone quotes, 2.1 million unfilled jobs by 2030, comes from the 2021 study. The 2024 follow-up put the 2024-to-2033 shortfall closer to 1.9 million and emphasized that the challenge includes both a skills gap and an applicant gap. Manufacturers aren’t only short on trained people. They’re short on applicants of any kind for open roles.
The 2.1 million figure everyone quotes comes from 2021. The newer study puts the need near 3.8 million through 2033, and roughly three-quarters of it is replacing retirees rather than staffing industry growth.
For maintenance and reliability teams, the cleaner signal isn’t the trillion-dollar headline. It’s the BLS occupation data: more than half a million industrial machinery mechanics, maintenance workers, and millwrights, growing 13 percent through 2034, with about 54,200 openings a year, many of them to replace people who move to other work or leave the labor force. That’s the maintenance hiring demand teams can plan around. It isn’t a count of unfilled jobs.
Why the numbers vary so much
Four reasons the figures disagree, and none of them mean the data is junk.
Study vintage. The Deloitte and Manufacturing Institute series has produced different headline numbers in 2015 (2 million by 2025), 2018 (2.4 million by 2028), 2021 (2.1 million by 2030), and 2024 (1.9 million by 2033). Each replaced the last. People quote whichever one they first heard.
Projection versus measurement. The shortfall numbers come from an economic model with assumptions about growth, retirement, and participation. Change the assumptions and the output moves. A BLS employment count doesn’t have that wobble, which is part of why it scores higher here.
Skills gap versus applicant gap. The 2024 study made this explicit. Some openings go unfilled because candidates lack the right skills. Others go unfilled because nobody applies at all. Older write-ups blur the two.
Retirements drive most of the need. In the 2024 study, 2.8 million of the 3.8 million projected openings come from retirements, roughly three-quarters, with 760,000 from industry growth and 230,000 from federal investment. That creates a particular risk for maintenance teams, where experience can take years to build and walks out the door at retirement.
How to use these numbers safely
Cite the 2024 study, not the 2021 one, when you need a current shortfall figure. If you reference 2.1 million, label it as the 2021 projection so a sharp reader doesn’t catch you using a stale number.
For anything maintenance-specific, lean on BLS occupation data rather than the whole-sector shortfall. “About 54,200 openings a year for industrial machinery mechanics, maintenance workers, and millwrights, many from workers transferring out or retiring” is defensible and specific. “2.1 million unfilled maintenance jobs” is neither.
Treat the median-age stat as a modest gap, not a cliff. Around 44 versus 42 is real and worth planning for. It is not the dramatic outlier some posts make it sound like, and the overall workforce is aging too.
Separate sentiment from vacancy. NAM survey results tell you manufacturers are worried. They don’t tell you how many seats are empty. Use each for what it measures.
These numbers don’t sit on their own. A thinner, greener crew tends to lean on reactive work, which feeds downtime cost and widens the planned-versus-reactive spend gap, while predictive maintenance can take some load off stretched techs. If you’re pressure-testing a maintenance budget against that risk, maintenance cost as a percent of RAV is the companion benchmark.
Where teams go wrong
The most common mistake is quoting 2.1 million as if it’s a live, present-day count. It’s a 2021 projection for 2030, and a newer study has already revised it.
Close behind is conflating the studies. You’ll see “4 million jobs, 2.1 million unfilled, by 2030” in a single sentence. The 4-ish million comes from 2018 (for 2018 to 2028), the 2.1 million from 2021 (for 2030). Smashing them together produces a number that exists in no study.
Then there’s the “10,000 baby boomers retire every day” line, dropped into maintenance hiring posts as if it describes the trades. It describes the entire U.S. population turning 65, across every industry. It says nothing specific about manufacturing maintenance, and Deloitte itself flagged the figure’s limits years ago.
Last is treating maintenance as if it has its own published shortage number. We didn’t find one in the sources reviewed. Anyone who hands you a precise “maintenance skills gap” figure is almost certainly repackaging a whole-sector projection. The honest move is to size it from BLS occupation data and say so.
Methodology
Each figure earns a tier based on three tests. Does it trace to a named primary source rather than a citation chain of blogs? Is that source current, or has it been superseded? And does the number measure what it’s being used to claim?
High means the figure comes from a primary source, is current, and is used for what it actually measures. Government employment data and the latest study in a series sit here. Medium means the figure is real but dated, modeled with shifting assumptions, or behind a wall that prevents full verification. Low means the sourcing is thin, the figure is extrapolated from a weak base, or it gets applied to manufacturing when it was never about manufacturing.
A negative finding can score High when it’s stated narrowly. “We found no authoritative national maintenance-specific skills-gap estimate in the sources reviewed” is a confident, defensible claim, even though it isn’t a number.
The Short Version
The manufacturing skills gap is well documented and the concern is durable, but the headline figures need handling.
Use the 2024 Deloitte and Manufacturing Institute study for a current shortfall: a net need for roughly 3.8 million workers through 2033, about 1.9 million potentially unfilled, with 2.8 million of the total (roughly three-quarters) coming from retirements. Retire the 2021 “2.1 million by 2030” line, or label it as the older projection it is.
For maintenance specifically, skip the whole-sector numbers and cite BLS: more than 538,000 industrial machinery mechanics, maintenance workers, and millwrights, growing 13 percent through 2034, about 54,200 openings a year. We found no separate, authoritative “maintenance skills gap” number in the sources reviewed, and that’s worth saying plainly.
Sources
- Deloitte and The Manufacturing Institute, “Taking charge: Manufacturers support growth with active workforce strategies” (2024 Talent Study), April 2024. https://www.deloitte.com/us/en/insights/industry/manufacturing-industrial-products/supporting-us-manufacturing-growth-amid-workforce-challenges.html Figure 2 (with the 2.8M/760k/230k breakdown) is in the report PDF: https://themanufacturinginstitute.org/wp-content/uploads/2024/04/Digital_Skills_Report_April_2024.pdf
- The Manufacturing Institute, “2.1 Million Manufacturing Jobs Could Go Unfilled by 2030” (2021 study summary). https://themanufacturinginstitute.org/2-1-million-manufacturing-jobs-could-go-unfilled-by-2030-11330/
- Deloitte and The Manufacturing Institute, “Understanding the skills gap in the manufacturing industry” (2018 skills gap and future of work study). https://www.deloitte.com/us/en/insights/industry/manufacturing-industrial-products/manufacturing-skills-gap-study.html
- National Association of Manufacturers, Manufacturers’ Outlook Survey, First Quarter 2024 (workforce as top concern). https://nam.org/2024-first-quarter-manufacturers-outlook-survey/
- National Association of Manufacturers, Manufacturers’ Outlook Survey, Q1 2026 and Q2 2026 (trade, healthcare, and input costs as the leading concerns). https://nam.org/2026-first-quarter-manufacturers-outlook-survey/ and https://nam.org/2026-second-quarter-manufacturers-outlook-survey/
- S. Bureau of Labor Statistics, Occupational Outlook Handbook: Industrial Machinery Mechanics, Machinery Maintenance Workers, and Millwrights (2024–2034 projections, May 2024 wages). https://www.bls.gov/ooh/installation-maintenance-and-repair/industrial-machinery-mechanics-and-maintenance-workers-and-millwrights.htm
- S. Bureau of Labor Statistics, Current Population Survey: Employed people by detailed industry and age, 2025 annual averages. https://www.bls.gov/cps/cpsaat18b.htm
- S. Census Bureau, “Manufacturing Faces Labor Shortage as Workforce Ages” (2020). https://www.census.gov/library/stories/2020/11/manufacturing-faces-labor-shortage-as-workforce-ages.html
- S. Census Bureau, “U.S. Workforce is Aging, Especially in Some Firms” (2025). https://www.census.gov/library/stories/2025/12/older-workers.html
- The Manufacturing Institute, “The Aging of the Manufacturing Workforce.” https://themanufacturinginstitute.org/research/the-aging-of-the-manufacturing-workforce/
- Panopto, “Workplace Knowledge and Productivity Report” (with YouGov, 2018). Cited as the origin of the “$47M” figure, not as an authoritative manufacturing source. https://www.panopto.com/resource/ebook/valuing-workplace-knowledge/
