This is Part 2 in our series on manufacturing in America. Part 1 asked why Americans aren't filling the manufacturing jobs that are already here. Subscribe to the Planet Money newsletter for the next installment. As always, our podcast is here.
Politicians in both big political parties have been trying to reengineer the U.S. economy to boost manufacturing. What makes manufacturing so special? Besides, you know, politics.
Some writers and economists say that romanticizing this one particular sector of the economy is more about nostalgia or political pandering than rational thinking about what will best serve workers and the economy.
Putting aside the debate about whether politicians can successfully bring back a large number of manufacturing jobs — many economists are skeptical of this — we at Planet Money were curious, on a deeper level, about what makes manufacturing special enough to warrant all this ruckus in the first place?
So we pored over nerdy econ studies and data, and we called up a bunch of economists, including two Nobel laureates.
We're going to look at this question from two perspectives. The first is whether manufacturing is special for workers. That will be the focus of this Planet Money newsletter. The second, which will be the focus of a forthcoming newsletter, is whether manufacturing is special for the broader economy.
Manufacturing pays a premium
"Manufacturing is special," says Gordon Hanson, an economist at Harvard Kennedy School who has published influential research on American manufacturing. " That's because as long as we've been able to measure earnings in the sector, it's just paid workers more, especially workers without a college education."
Economists call this "the manufacturing premium." And it's worth noting that some research suggests that the manufacturing premium has fallen or even disappeared in recent decades.
However, Hanson and other economists we spoke to said the most convincing evidence on this subject shows that the manufacturing premium is still alive and well. In particular, they cited a recent, peer-reviewed study from economists David Card, Jesse Rothstein and Moises Yi. Hanson calls it "the gold standard." The economists use much richer, more comprehensive data than previous studies. And they deploy this pretty cool technique that helps them see, systematically, what happened to the pay of over 100 million Americans as they jumped between industries.
" We basically take every single workplace in the United States, and we watch workers as they move from workplace to workplace," says Card, an economist at the University of California, Berkeley who — ummm, flex — won the 2021 Nobel Memorial Prize in Economic Sciences. (Read this Planet Money newsletter that explains why he won the Nobel.)
Now, it's no secret that industries like tech, finance, law and medicine pay workers a lot. But those high-earners usually have advanced degrees and valuable skills or abilities. Had they not gone into their chosen industry, they would probably have made good money in another industry. It's hard to know how much of their pay is the result of their own special sauce and how much of it is driven by something else in their chosen industry. Like how profitable it is. Where it's located. How unionized it is. The management or culture or size of its companies. Whether it's competitive. And so on.
It's a similar issue when analyzing manufacturing. The average manufacturing worker makes over $35 an hour, according to the Bureau of Labor Statistics. That's a lot more than the average earnings of workers at restaurants, which is about $21 an hour. What accounts for this pay difference? The workers themselves or something about these industries?
Imagine what would happen if you could run an experiment and you could see, systematically, what happened when you plopped restaurant workers into manufacturing, or vice versa. That's kinda what the economists do in this paper, statistically speaking. And it's not just restaurants and manufacturers. It's basically every industry.
Using a massive dataset, the economists track what happens to millions and millions of workers as those same workers move to different industries. This allows them to control for age, education and experience, like other studies do, but they can also go a step further. Because they're tracking the same workers, they're controlling for worker qualities that are not easily seen in data — like their grit, their know-how, their work ethic, their rizz.
With this technique, the economists can cut through more of the noise of why particular industries pay particular workers particular wages. They can get more to the heart of the actual, causal effect on what working in a particular industry can do to people's incomes — in a sense, how special an industry is.
Card, Rothstein and Yi use restaurants as the base industry to compare with other industries. This, they write, is the industry with the lowest average earnings in their data. Think fast-food and cafeteria workers, cooks and servers. The baseline could really be any industry. It's a bit arbitrary because they measure all industry premiums against this same baseline.
You can think of the industry premiums they find below as basically the pay bump the average person would get from quitting a job in this mostly low-wage service sector and getting one in another industry, without going back to school or dramatically changing the skills or capabilities they already have.
The manufacturing premium vs. other industries
The economists calculate that, on average, if a worker jumps from working in restaurants to working in manufacturing, they get a 35% pay bump.
That's a substantial premium, higher than if they went into retail (an 11% premium), education (a 13% premium), agriculture (a 16% premium), health care (a 19% premium), transportation and warehousing (a 24% premium), construction (a 25% premium), finance and insurance (a 32% premium) and professional, scientific and technical services (a 33% premium). Pretty special!
However, the manufacturing premium is lower than that of some other sectors, including "information," which includes the movie, publishing, broadcasting and telecommunications industries (a 39% premium); utilities (a 49% premium); and mining, quarrying, oil and gas (a 62% premium). OK, so manufacturing is medium-ish special.
That said, the 35% percent pay bump in manufacturing is the average for a pretty diverse sector. Manufacturers make all sorts of things: socks, rockets, potato chips, petrochemicals, mannequins, skateboards, electronics, cars, Twinkies, you name it.
So Card, Rothstein and Yi break down the manufacturing premium, looking at what it is in various subsectors.
They find that in the low end of the manufacturing spectrum are industries that make apparel and leather (a 15% premium), furniture (a 19% premium) and textile mill and wood products (a 22% premium). Hmmm. Not so special compared with a lot of other industries.
In the middle are manufacturing subsectors like plastics and rubber (a 30% premium) and food products (a 32% premium). OK, yeah, now we're cooking.
And at the high end are industries that manufacture computers and electronics (a 42% premium), transportation equipment, so like cars, ships and planes (a 43% premium), chemicals (a 47% premium) and — the highest of all — petroleum and coal (a 62% premium).
Wowzer. That last one especially — that's a big premium. In fact, next to mining, it has the highest premium in all the industries the economists look at. It helps explain why — despite rising concerns about climate change — these fossil fuel-producing jobs are so politically popular in many parts of the United States.
A clear takeaway from this data is that high-end manufacturing offers workers a significant pay bump relative to other industries. Low-end manufacturing, of things like apparel and furniture, not so much.
Why does manufacturing pay a premium?
Card, Rothstein and Yi don't provide definitive answers for why manufacturing pays a premium. There are many plausible explanations.
One is that manufacturing has a history of being a more unionized sector, which led to higher pay and benefits. A core idea here is unions force business owners and managers to share more of their profits with workers.
Daron Acemoglu, the 2024 winner of Nobel Memorial Prize in Economic Sciences, told us he believes manufacturing is special because, at least historically, it has been a union bastion.
" Manufacturing is super important for unionization," Acemoglu says. "So if you want to have unions in your economy, you want manufacturing. They unionize themselves and they unionize other sectors."
Some pro-union advocates have gone so far as to argue that Americans are nostalgic for an era of lots of manufacturing jobs not because they were good jobs per se, but because they were good, union jobs. The implication: Promote unions, not manufacturing.
Economists debate why manufacturing has been more prone to unionization, but one leading explanation is the fact that workers in this industry tend to work at one large workplace (the factory) and, at least during some periods historically, they had lower turnover, which maybe helps create more personal connections and solidarity between workers. Plus, under U.S. law, unionization has to happen at the workplace (or "establishment") level, and unionizing smaller service-sector workplaces is harder. Instead of unionizing one big factory, workers have to unionize Starbucks by Starbucks or Walmart by Walmart. (Read more about unions in a previous Planet Money newsletter.)
However, it's also true that the unionization rate in manufacturing has fallen pretty precipitously in the United States and that many manufacturing plants are not unionized, especially in the Midwest and South, where a lot of states have passed laws that make unionizing more difficult.
In fact, in 2023, only about 7.9% of manufacturing workers were unionized, which is not crazy different from the overall union membership rate in the private sector, which stood at about 5.9%.
Another potential hypothesis for why there's a manufacturing premium: Working in factories may be more dangerous than working in a lot of service jobs, and so manufacturers may have to pay a premium to attract workers willing to take on greater risk (economists call this a "compensating differential"). This is a classic explanation for why coal mining pays a premium.
However, the data doesn't seem to really support this hypothesis in manufacturing. According to data from the Bureau of Labor Statistics, in 2023 the fatal injury rate in manufacturing was 2.5 per 100,000 workers. That's slightly higher than industries like restaurants (1.6), retail trade (2.1) and leisure and hospitality (2.3) but well under construction (9.6), transportation and warehousing (12.9) and agriculture, forestry, fishing and hunting (20.3). Manufacturing also doesn't look particularly more dangerous than other sectors when it comes to nonfatal injuries or illnesses.
It's also possible manufacturers have to compensate workers because many believe factory work is unpleasant. In fact, several large Reddit threads formed in response to the last Planet Money newsletter on workforce issues in manufacturing. Many commenters were manufacturing workers themselves. And some talked about how uncomfortable factory jobs can be — including details like factory floors can get really hot when you're wearing protective gear and standing amid a bunch of machines humming and whirring — how shift schedules can be long and inflexible and how the jobs can be mentally or physically taxing.
The Pew Research Center recently did a survey that suggested that manufacturing workers are among a more general group of "blue-collar" workers who are more dissatisfied with their jobs than the rest of the workforce. This group also includes those who do manual or physical labor in sectors like mining, agriculture, retail, transportation, construction and restaurants.
A 2024 Gallup survey found that only 23% of manufacturing workers felt engaged or enthusiastic about their jobs. That was lower than the overall workforce (31%). But it was also lower than some other blue-collar industries, including construction (32%), utilities (30%) and agriculture and mining (29%).
Dislike of factory work might help explain why, despite manufacturing paying a premium, many workers aren't filling these jobs — which may actually increase the size of the premium that manufacturers have to pay to find people to do them.
The last big reason that manufacturing may pay a premium: Manufacturing workers handle a lot of expensive machinery. Most of the economists we spoke to stressed this factor. It helps explain the pattern in the premium data we went through before, where more advanced manufacturing — which is filled with a lot of expensive machines — pays an even higher premium.
For one, employers may have to pay more to attract people whom they can trust will be careful with their expensive machinery and not screw up productivity on the assembly line. The machines that workers have to operate, Card says, are oftentimes very expensive, "so you don't want someone goofing off."
Or, worse yet, you don't want workers to intentionally damage this expensive equipment. Rothstein, a co-author of this study and another economist at UC Berkeley, says this fear was more relevant back in the day, when there were violent industrial conflicts. Today, probably not as much. But, he says, it's still true that "the more capital-intensive you are, the more damage workers can do to you if they decide to go to war against you." Don't get these workers angry!
Probably the most convincing factor though: There's a lot of learning on the job to figure out how to work with this expensive machinery. We reported in our last newsletter how workers with machine-related skills are very needed in the industry.
Teaching these skills takes time and money. So maybe employers pay extra to convince workers to not quit and to lower their turnover rates. They also may use higher wages to motivate employees to work hard and be diligent because it's hard to monitor huge factory floors (this is known in economics as the "efficiency wage hypothesis," a theory that employers pay extra to accomplish workforce goals like these).
More generally, working with machinery often means you're productive, and productivity is one important factor influencing pay. Also, these advanced manufacturers with expensive machinery are often big, profitable companies.
" The kind of standard way that economists have always thought about this is that big firms and capital-intensive production are both good for wages," Rothstein says. "And that's manufacturing, especially advanced manufacturing,"
Whatever the reason, according to this research, manufacturing does continue to pay a premium and especially so in advanced manufacturing, where there's a lot of expensive equipment. Pretty special!
Manufacturing punches above its weight when it comes to providing good jobs
Gordon Hanson, together with UC Berkeley economist Enrico Moretti, recently did a study in which they use a measure of industry pay premiums, similar to Card, Rothstein and Yi, to see which industries have provided good jobs to Americans over the last five decades.
Now, their definition of "good jobs" isn't really how we typically talk about good jobs. Like Card, Rothstein and Yi, they control for worker characteristics when calculating which industries provide jobs with a high pay premium. That takes the focus away from higher education and credentialed skills as the tickets to a well-paying job, because it's holding things like that constant. It's no secret higher education pays in today's job market.
In this paper, what they're really interested in, Hanson says, is "the job market lottery." That is, if you don't necessarily have a college degree and if market forces in the job market are kinda stacked against you, which industries allow you to win the lottery and reach the middle class? They define good jobs as those that pay high premiums (in the top third), meaning there's something about these industries that gives workers a big extra pay bump regardless of their background.
With this definition of good jobs, they find that, back in 1980, manufacturing accounted for almost 40% of all good jobs in America. That was for both college- and non-college-educated workers. It's a pretty stunning statistic.
The economists find — not surprisingly — that manufacturing's portion of the nation's good jobs has fallen dramatically over the past few decades as those jobs were automated and offshored away.
Today, manufacturing accounts for only about 10% of the American workforce. Yet, with this measure, it accounts for about 20% of all good jobs in America. Manufacturing punches far above its weight when it comes to industries that provide an extra pay bump to Americans.
So, yeah, Hanson says, manufacturing jobs are special. But over the last few decades many have disappeared. And opportunities for non-college-educated Americans — who represent more than half the workforce — shriveled. The one bright spot, the economists find, was in construction and the skilled trades.
Meanwhile, good jobs in industries like finance, tech, management and legal services blossomed. These sectors are weird though, Hanson says. They provide tons of opportunities to workers with a college education but very few to those without one.
" There are certain types of industries where you can't unlock the productivity magic box without a college degree," Hanson says. " In manufacturing, you can open the magic box without a college degree."
That said, as we reported in our last newsletter, many manufacturing jobs these days also require a bachelor's degree, and even ones that don't often require an associate degree, a certificate, an apprenticeship or critical vocational skills. Yet, still today, unlike in other sectors, these business enterprises are often socioeconomically and educationally diverse. Despite having higher entry barriers than in past generations, manufacturing still has lower entry barriers for non-college-educated Americans than many other industries to get in the door and see higher wages. And many stress that this is particularly true for non-college-educated men, a population that has seen economic struggles in recent decades.
Does that mean the government should do everything it can to boost this one sector?
Many economists doubt whether the government can do much to dramatically boost the number of manufacturing jobs in the country. For one, many of these jobs have become automated with machines in recent years, and fewer workers are needed. Tariffs might help bring some of these jobs back. But tariffs and the trade retaliation they invite could also hurt other areas of the economy that also provide good jobs — including in American manufacturing, where companies often need foreign parts and materials to efficiently make things in America.
Tariffs also raise prices for everyone, including working-class Americans who make less income and will feel it the most.
"I think we've developed a kind of collective fetish for manufacturing, which is really unproductive," Hanson says. "The problem is not too few manufacturing jobs. The problem is too few jobs for workers without a college education."
Hanson says, sure, we should try to create more good jobs in manufacturing. He — like many other economists — is skeptical whether we can bring back labor-intensive manufacturing, like in apparel and furniture. And as we showed before, those pay low premiums anyways. They're not a great source of high-paying jobs. And, Hanson says, these jobs are likely not going to come back, because cheap labor is really important to their business model. Slapping tariffs on this stuff, Hanson says, will mean "we're just gonna charge U.S. consumers a whole lot more for the privilege of consuming T-shirts and tennis shoes and board games and the like."
America continues to have a high-end manufacturing sector that produces things like cars, planes and petrochemicals. And Hanson says we may "have a shot" at bringing back some manufacturing jobs in the middle, which produce things like auto parts and machine tools. He says he's open to the idea that strategic, temporary tariffs can help make that happen. He, however, thinks it's much more important for America to invest in workforce development programs, like reforming high school education to offer kids more routes to get vocational skills, and investing in community college programs and apprenticeships to equip Americans with the skills needed to be productive in manufacturing.
But the reality, Hanson says, is that manufacturing is only a small slice of the economy — and it will always be a small slice of the economy. So, he says, leaders should be focusing a lot more on creating good jobs in services, because "most of the good jobs are going to have to be in services."
And that could mean trying to replicate some of the things that have made manufacturing jobs special. Like potentially reinvigorating unions, or trying to get more high-paying industries with lots of advanced technology and machinery to open their doors to people without a college education. Hanson imagines companies and the government figuring out more ways to create good jobs for non-college-educated folks in sectors like tech or health care. Our education system and emerging technology could help make that happen.
Manufacturing, however, is about more than just good jobs. Coming up in the Planet Money newsletter: We look at why manufacturing may or may not be special for the broader economy.
If this was forwarded to you or you're reading online, you can sign up here.
Copyright 2025 NPR
