265 | John Skrentny on How the Economy Mistreats STEM Workers
What this covers
John Skrentny, a sociologist at UC San Diego, sits down with Sean Carroll to discuss his book Wasted Education and a central paradox: American employers constantly claim a dire shortage of STEM workers, yet only 28–40% of STEM graduates actually work in STEM jobs. Rather than a shortage, the data suggests an oversupply. Skrentny argues that the contradiction dissolves once you look at how the modern American economy treats STEM workers—not as assets to develop, but as disposable labor to be burned through quickly and replaced with cheaper newcomers. The conversation traces this pattern to a deeper structural shift: the rise of maximizing short-term shareholder value since the 1980s, which recast workers as costs to be shed rather than as people worth retraining.
The discussion moves across several connected territories. Skrentny lays out how employers exploit STEM workers' genuine passion for their fields, coaxing them into crunch mode—60–70 hour weeks without extra pay—until they burn out, then replace them with fresh graduates. He contrasts this with professional sports, where scarce talent commands premium pay, whereas STEM managers typically earn more than the technical workers whose skills are supposedly critical. The economy diverts capital toward software—quick to demonstrate, high-margin—while starving hardware innovation in clean energy and fusion that society desperately needs. Skrentny also examines the gold rush cycle of tech booms: blockchain, then AI, then the next thing, with workers perpetually chasing new fields as old ones cool. He proposes partial remedies within existing capitalism—government incentives like the Inflation Reduction Act, moral constraints on harmful STEM uses, and a shift toward lifelong, periodic retraining rather than one education ending in a worker's early twenties.
Skrentny argues that despite constant claims of STEM worker shortages, the US economy systematically mistreats STEM graduates—burning through them as cheap, disposable labor with rapidly obsolescing skills—because a short-term-shareholder-value version of capitalism treats workers as costs rather than assets to be retrained.
- Only ~28-40% of STEM graduates actually work in STEM jobs, contradicting the shortage narrative
- Employers 'burn and churn' workers, refusing to invest in retraining and preferring cheaper new graduates
- A 'maximizing short-term shareholder value' culture rewards shedding workers and favors software over the hardware innovations society needs
Tech companies systematically exploit worker passion and enthusiasm until burnout and obsolescence.
- New sociological research shows employers exploit STEM workers' passion for their field—coaxing them to work 60-70 hours a week without extra pay because their enthusiasm, instilled by good teachers and professors, makes them willing to overwork until they burn out.
“there's new research, mainly from my field and sociology, about how employers exploit this passion. They think, "Wow, we've got these workers who really love this and let's get 'em to work 60, 70 hours a week. We don't even have to pay them extra 'cause they're so passionate."”
- An ethnography by sociologist Ofer Sharone found tech workers averaging 67 hours a week during 'crunch mode,' voluntarily overworking because forced/stacked ranking ('rank and yank') and a desire to impress managers lead them to overpromise on ambitious deadlines—rendering the foosball and slides of fun workplaces largely unused.
“He found workers worked an average of 67 hours a week... And they did it voluntarily.”
- STEM work tends to be a young person's game because keeping skills constantly updated is hard and burns people out, and because older workers with families have a 'second shift' at home that makes self-training in one's own time increasingly difficult and stressful.
“it's often a young person's game... older workers often have families, and if you have families, they take time... it gets harder and harder to self train in your own time.”
- STEM workers must continuously update their skills throughout their careers just to keep their jobs—the 'STEM skills treadmill'—meaning education does not end at college or even a PhD, and failure to keep re-skilling leads to obsolescence and exit from the field.
“I call it the STEM skills treadmill... I really wanted to capture something here, which is the idea of moving while standing in place. STEM workers really have to keep moving. They really have to keep learning just to be able to keep their job.”
- STEM work resembles 19th-century prospecting and mining—workers chase successive technology 'gold rushes' (blockchain, then AI, then the next thing) where a field gets hot, hires heavily, and some get rich before it peters out, leaving workers to chase the next boom; one engineer found this so exhausting he discouraged his children from entering these fields.
“he thought STEM work was something like prospecting and mining in the 1800s... it gets tiring chasing these gold rushes... AI is now the huge one. It goes up and down.”
STEM labor surplus and corporate contradictions expose the shortage narrative as false.
- Rich, near-monopoly tech firms (Google, Amazon, Meta, Apple) fund a lobbying group, FWD.us, whose reason for being is the STEM shortage narrative—lobbying for more STEM education and migrants—even as they conduct repeated mass layoffs, exposing the contradiction between their shortage claims and their behavior.
“They created this lobbying group called Forward US, that their whole reason for being is the idea of a STEM shortage... A lot of these companies... they're rich. Some of these are near monopolies.”
- Employers loudly claim a dire shortage of STEM graduates, yet only about 28% (Census Bureau) to 40% (Skrentny's counting) of STEM graduates actually work in STEM jobs, suggesting the data shows a surplus rather than a shortage—the central puzzle of the book.
“According to some measures, and this is the Census Bureau, only 28% of STEM grads work in STEM jobs... My counting is about 40% of STEM grads work in STEM jobs. So it's a little bit higher, but still a minority.”
- Innovation-economy employers deliberately 'burn and churn' STEM workers—exploiting their enthusiasm while their skills are current, then replacing them with fresh, cheaper graduates once they burn out or their skills become outdated—treating workers as disposable rather than investing in retraining.
“Yeah, we burn and churn through them. They have a lot of enthusiasm and we take advantage of that. And when their skills fall out of date or when they get burned out, then we hire some more.”
- The shift in American capitalism toward 'maximizing short-term shareholder value' since the 1980s redefined corporations around shareholders rather than employees or customers, recast workers as costs to be shed, and led investors to applaud layoffs even in good economic times—pushing the stock price up when workers are cut.
“by the 1980s as we got to shift to this maximizing short-term shareholder value approach to corporations, you got companies that shed workers even in good times. And the stock price would go up.”
Education must shift toward lifelong re-skilling and foundational thinking rather than narrow technical training.
- College education's enduring value is creating foundational skills—especially learning how to learn and discerning good knowledge from bad—which remain valuable even as specific deployed skills change many times over a career, so degrees should be paired with broad communication and critical-thinking skills rather than narrow technical training alone.
“one of the things is to learn how to learn. PhD training really makes you good at that, but even with the bachelors, you kinda learn how to learn. You can discern good knowledge from bad knowledge”
- Because technical skills regularly become obsolete (illustrated by Apple dropping Adobe Flash, instantly stranding Flash experts), the traditional model of education ending in one's early 20s must change toward lifelong learning—returning to college periodically for short, specific re-skilling certificates rather than full new degrees, possibly via a subscription model.
“we need a model where we think of college as a place you return to periodically for skills updating or re-skilling... Maybe something like a subscription model”
- Rather than radical change, the existing tools can extract more value from STEM investments: government incentives (like the Inflation Reduction Act) for clean energy and needed innovations, legally banning harmful uses of STEM skills such as targeted advertising, and applying moral limits to investor behavior the way society already constrains cannabis or pornography investment.
“I actually don't think we need to think more radically about it. I think that we have the tools available now to get more out of our investments in STEM education than we are getting now.”
- Sociology produces 'horizontally integrated' knowledge—broad but more superficial across many connected domains—which serves as a corrective to economists' 'vertically integrated' tunnel vision, as when economists' basic utilitarianism failed to anticipate that free trade's job dislocation would harm people whose identities are bound to their work.
“I think of the science that you do, Sean, is like vertically integrated... Whereas the stuff that I do... it is horizontally integrated knowledge. It's a lot of stuff at a more superficial level, but really broad.”
- The US government does not classify practicing medicine as a STEM job because STEM policy is oriented toward creating innovation, patents, and jobs, whereas practicing physicians mostly deploy existing expertise rather than innovate.
“the government does not consider practicing medicine to be a STEM job... It's because of this thing about trying to create innovation... for the most part, practicing medicine is about deploying existing expertise.”
- When firms outsource roles like user experience design to cheaper workers abroad (Eastern Europe, China, Mexico, India) rather than pay the going domestic rate, it both reflects isomorphic wage-setting and reduces the incentive for Americans to enter that field, making the economy more dependent on outsourcing over time.
“he went around the world... looking for cheaper user experience designers... the more that kind of thing takes off, the less you incentivize people to move into that field in the United States, the more you become dependent on outsourcing that thing.”
- The STEM acronym was coined by Judith Ramaley, a National Science Foundation official, who reordered earlier failed acronyms like SMET and SET; it then took off in congressional bills and spread internationally.
“a, Judith Ramaley is her name, she was one of the higher ups at the National Science Foundation. She apparently is the one who coined the STEM acronym.”
Software's porosity and stack-ranking pay structures create perverse incentives driving workers toward management.
- Software development is unusually porous compared to electrical engineering or life sciences—about 15% of STEM workers lack STEM degrees, and people from humanities backgrounds can enter through bootcamps in coding or data analytics, making it welcoming to non-traditional entrants.
“a good 15% of STEM workers don't have STEM degrees... One of the great things about software development is it's a field that's pretty porous... software is wonderfully porous.”
- Unlike professional sports where players with scarce skills are paid more than managers, in STEM the managers are typically paid more than the workers whose skills are supposedly scarce, which incentivizes workers to leave technical roles for management—a self-inflicted wound that worsens any genuine scarcity.
“in professional sports, the players who have these scarce skills are paid more than the managers. In STEM, the managers are typically, not always, typically paid more than the folks who have the skills that are supposedly scarce.”
- Cash-rich tech companies that complain of shortages could adopt an Individualized Education Plan (IEP) model from special education—monitoring workers' strengths, anticipating which skills will become obsolete, and proactively funding retraining to redeploy them—but they typically refuse, instead giving workers a few months to figure out a new skill entirely on their own.
“Google... basically told employees, "You've got this many months to figure out what skill you want to develop. You're on your own to figure that out." How do you know what's going to be the next in demand skill?”
Workers suffer moral injury and job dissatisfaction when STEM skills are applied to ethically questionable ends.
- Some STEM workers experience moral stress or moral injury because they are uncomfortable with their firm's business model—Facebook workers describing 'hurting people at scale' and feeling they 'have blood on my hands' for the platform's role in disinformation and political conflict, with evidence linking social media to rising teen depression, especially in girls.
“I quote some workers at Facebook who've said Facebook is hurting people at scale. Another worker more graphically said, "I feel like I have blood on my hands for Facebook's role in promoting disinformation and promoting political conflict in societies other than the United States."”
- Platforms like YouTube use recommendation algorithms that lead users toward more extreme, shocking, or emotionally provocative content because getting an emotional hook by outraging users—'engagement equals enragement'—keeps them using the app longer.
“There's a phrase "engagement equals enragement"... the idea is getting this emotional hook into the users by out-raging them is a way to get them to use the app longer.”
- Research using the National Survey of College Graduates found that STEM-degree holders who went into non-STEM jobs were significantly less likely to report being satisfied with their jobs overall, and notably less satisfied with the intellectual challenge of their work, suggesting they are more likely to be bored.
“we looked at people with STEM degrees who went into non-STEM jobs, is they were significantly less likely to report being satisfied with their jobs, overall. And, interestingly, they were significantly less likely to be satisfied with the intellectual challenge of their jobs.”
Investment preferences and government policy systematically misdirect capital toward software over needed hardware innovation.
- Investors strongly prefer software because it can be quickly demonstrated to work, requires no factory or large workforce, and yields returns roughly ten times higher than clean energy—diverting capital away from the hardware innovations society desperately needs like clean energy, better plastics, and fusion, which face the 'valley of death' between innovation and manufacturing.
“The return on software investments are 10 times, much percentage higher than returns on clean energy, which is what I think we badly need.”
- Cold War-era and modern government STEM policy operates on a quasi-religious 'faith'—what Skrentny calls 'alchemy'—that simply creating more STEM-educated people will produce amazing, unspecified innovations, without the government knowing which fields actually generate the most jobs, competitiveness, or innovation.
“by the Cold War, we started to get into a phrase that I called it alchemy, 'cause it suggested this hope that by putting stuff together, something amazing would happen. It wasn't as rational”
- About 50% of all STEM workers in the United States are in computer-related fields, making computing the '800-pound gorilla' of STEM; unlike chemists who are hired only by chemical firms, computer experts form a band across the entire economy, hired by trucking, retail, life sciences, and every sector.
“about 50% of all STEM workers in the United States are in computers in some way. That's like the 800-pound gorilla of STEM.”