I Used to Believe in STEM

The Foibles of STEM Boosterism

Photo by Chromatograph on Unsplash

How long has the STEM craze been going on—ten or fifteen years, maybe? Back then I was an unhesitating devotee. I thought it was all pretty obvious: STEM both pays better and it’s more useful, and I didn’t buy counterarguments about the supposed value of most non-STEM degrees. Upon reflection, though, my cheerleading for STEM was probably more tied up with residual animus towards the humanities than any well-evidenced convictions about the utility of STEM education. I distrusted commonplace reassurances about the subtle epistemic virtues of the liberal arts, which were all perpetually falsified by largely tepid employment prospects and wimpy salaries. And I resented continually having to study things that I didn’t respect.

I particularly disliked the repeated attempts to re-brand STEM to "STEAM" to reincorporate the very stuff people were trying to implicitly discourage by supporting STEM. It seemed like a refusal to be honest/frank about the situation. STEM, many of us assumed, was crucially unlike something like literature. I’m not wholly recanting my pessimism about the liberal arts, but the underlying presumption is that STEM is so markedly different was mistaken. Once you begin scrutinizing why STEM majors actually enjoy superior career outcomes, the whole pro-STEM worldview starts crumbling.

STEM Isn’t Really That Useful

The infatuation with STEM wasn’t just about pay scale—it was intertwined with a notion that it’s more useful—that the pay disparity is warranted and, unlike other subjects, STEM education augments students’ capabilities in a way that employers value by teaching them knowledge and techniques that are commonly used in private-sector work. But that’s mostly a fantasy.

I’ve mentioned before that much of the income disparity between college graduates and their non-college counterparts is attributable to signaling (indicating to employers that you’re smart, industrious, conformist, etc.) rather than universities furnishing students with employable skills (

Bryan Caplan

estimates it’s something like 80% signaling). But doesn’t that percentage vary across majors? Surely, medical education, for example, has to be primarily concerned with producing capable professionals rather than mere signaling, right? Isn’t the reason STEM folks get paid more because they also acquired practical knowledge and abilities in addition to the signal boost from graduating? Well, not really.

As with the wage premium for college graduates, a non-trivial portion of the wage premium for STEM graduates (over non-STEM) is probably also just signaling. Just as being a college graduate signals to potential employers that you’re relatively intelligent and studious, so does being a STEM major, and STEM is especially well-suited for signaling because it’s both arduous and amendable to objective assessment. I’m willing to believe that a smaller percentage of the wage premium for STEM is signaling-related than the wage premium for history grads, for example, but it’s much closer than people think.

credit: peralsofrawnerdism

I’m not happy about this, and neither are the underutilized STEM grads, but it’s the sad reality for most. The notion that STEM learning is regularly being deployed in the workplace by anything more than a pitiful fraction of the STEM workforce is an unduly romantic picture of the labor market. Truly cerebral work is a very recherché privilege that people are desperately competing for, and those with undergraduate degrees in biology, chemistry, or math are not often getting hired to make good money as any kind of biologist, chemist, or mathematician. Even something like computer science, which is supposed to be a paragon of marketable, skill-conferring education, is a richly ironic example, since there’s a sizable mismatch between CS coursework in universities and the job duties of most software devs.

Professional degree programs such as law or medicine are probably near the upper limit of how directly education can translate to workplace competency, and even those cases are surprisingly dubious. Ask your doctor friends about how competent the interns are when they show up for residency or how smoothly the teaching hospitals operate in July when the new cohort arrives; the bulk of a physician’s capabilities are acquired during their residency—on the job, in the hospital, focusing on their specialty—not in a classroom. Again, it’s not that all education is bogus, but if even these more tailored, professional-degree programs are really outputting malleable talent more often than plug-and-play workers, then presumptions about the utility of STEM education deserve heightened skepticism.

The reality is that even among the luckier share of talent/education, students are absorbing debt, foregoing earnings, and burning up cognitive resources learning about Maxwell’s equations and chirality to, ultimately, just dick around with spreadsheets in Excel for sixty hours a week or learn Next.js. To be clear the non-STEM grads are even worse off, but it’s not because those Excel jobs require a bunch of scientific background or familiarity with differential geometry or whatever.

STEM apologists, confronted with this prima facie impracticality, might suggest that laboring through stuff like biochemistry or quantum mechanics is indirectly beneficial because it equips the students’ minds with a general capacity to grapple with complex topics and problems, but that’s just the kind of nebulous, teach-you-how-to-think baloney that the humanities pushers constantly promise! Some workers really are hired for their skills and/or expertise, and I’d bet those people are happier with their work and enjoy pretty stable employment, but most of those skills aren’t the byproduct of university education—they’re likely the result of job experience and specialization.

STEM is a Faulty Category and Isn’t Exceptionally Lucrative

Another issue with promoting STEM is its inaccuracy. If we’re relying on career prospects to adjudicate what’s useful/needed, then STEM is a strange grouping. Good luck navigating the modern job market with an undergraduate biology degree from a lower-tier school: you’d likely be in better shape with a business degree or even a liberal arts degree from a more elite institution. And what about stuff like pharmacy, nursing, or accounting—I wouldn’t classify these as STEM, but there’s relatively good pay and involve learning real skills—probably more than most STEM degrees. There’s a nice thematic coherence to STEM, but labor-market demand isn’t aiming for thematic coherence, so the terminology is ultimately both over- and under-inclusive.

I assume STEM is meant to be restricted to the physical sciences (psych majors’ career prospects are abysmal). But if you, for example, blame students struggling to repay their loans for not taking a degree in one of these fields, that entails an extremely restrictive view of education:

Should psychology, history, political science, or even biology be off the table for prudential students? Exactly how many majors would be left unscathed? It’s too easy to joke about people majoring in Persian candlemaking or whatever when they could be studying finance; I rarely see these anti-forgivers bite any of the sizable bullets required to justify their stance. If they truly think stuff like psych and history should whither away, then they need to own that.

Additionally, undergraduate degrees simply aren’t remarkable anymore. If you’re aiming at something like the American Dream, with a reasonable expectation of enough income to pay for your student loans, childcare, and a house in a decent neighborhood, then the most important payout of an undergraduate degree is that it qualifies you for an particular subset of worthwhile graduate/professional programs, and your choice of major just doesn’t matter as much. Why not major in philosophy if you’re going to get a high-end MBA anyway? A lot of STEM majors wind up in the same place.

For most high earners, their economic advantage didn’t come from deciding on an undergraduate degree in chemistry over one in history. If that history major matriculates at the University of Chicago for a JD or MBA, they’d likely (massively) out-earn their STEM counterparts, and having a sociology degree doesn’t hamper your earnings as a physician. Fretting over majors has become an outmoded pastime, and it could even be a hindrance, given that worse grades from challenging coursework can prevent you from obtaining the optimal graduate degree.

The point is that there are multiple ways to swerve back into an employable trajectory following a liberal arts undergrad, and many of them are even more remunerative than STEM. If I knew someone who was talented enough to survive a STEM degree, then I might recommend that they do the same thing anyway. I wouldn’t try to dissuade them from STEM if they liked it, and maybe a segment of STEM grads (like computer science students ten years ago) don’t need that graduate education, but I doubt STEM is as economically necessary (or sufficient) as many pretend.

The other reason STEM majors earn higher incomes is that they’re simply more intelligent, conscientious, or care more about earning money. And it’s erroneous to think that promoting STEM will transmogrify the liberal arts folks into people with these same traits or somehow alter their underlying talents. Plus, is majoring in physics really such a promising idea for people who lack the the intellectual talent required for that degree? Or will it just ruin their GPA, injure their self-esteem, and delay their graduation? In truth, a lot of these non-STEM people would be pursuing STEM if they could, but they’ve already detected it’s a bad fit despite the well-known financial incentives for them to pretend otherwise.

STEM Is Anti-Intellectual

It’s also important to question whether pursuing STEM has drawbacks. The degree requirements for these subjects are usually pretty lengthy and will crowd out other studies. It’s a different mixture of classes and a different educational experience from the liberal arts, and that’s fitting for many STEM grads, but there could be some opportunity cost. Students are more driven to study whatever is economically advantageous these days, but people used to pursue education for enlightenment, and there are some leftover notions that universities provide this. The unhesitating celebration of STEM presupposes that the economic and non-economic aims of higher education coincide perfectly, as if students don’t really give up anything by studying STEM and organic chemistry and philosophy are equally intellectually enriching. They aren’t.

For all the exalting of how cerebral STEM majors are, the rigor and burden of those programs don’t necessarily translate to breathtaking contemplation. Learning worthwhile things is often difficult, but something being mentally taxing isn’t a surefire sign that it’s edifying, and grueling work isn’t invariably thought-provoking. Much of STEM education focuses on learning procedures for answering recognizable problems, like using Laplace Transforms to solve differential equations or even just memorizing stuff like the Krebs Cycle.

Sean Carroll has discussed the physics community’s longstanding hostility to foundational questions multiple times; David Albert even said that he nearly got ejected from his PhD program for his interest in such topics, and he was assigned a calculation-intensive dissertation as a corrective. This pervasive anti-philosophical stance towards the Copenhagen Interpretation of Quantum Mechanics is frequently characterized by the slogan “shut up and calculate.” People like Carroll are trying to reform the profession, but it illustrates the dangers of elevating STEM too much—it invites conformity and prizes verifiable answers over good questions. In short, it’s anti-intellectual.

Promoting STEM Isn’t Good for Society

Beyond the supposed benefit of STEM for the students, you might argue that promoting STEM benefits society. STEM graduates earn more money, so they’re likely more productive; hence, wouldn’t shoveling more people into STEM boost productivity and well-being? And, even if non-STEM students are eventually coerced into being useful to earn income, isn’t the welfare of society mainly advanced by technological improvements that are attributable to STEM?

This again makes the error of thinking that STEM education is reliably giving people particularized skills that they use at their jobs—that’s mostly a fiction, or at least less real than people think. And again, if STEM folks are more economically productive, that’s partly due to their natural abilities and dispositions, and forcing would-be liberal arts students into STEM programs won’t re-calibrate their genetic makeup.

On the issue of scientific and technological advancement, with the possible exception of AI, human innovation is slowing down, and it’s not like this is some policy misfire preventing companies from leveraging the STEM grads appropriately. Times have changed from the midcentury optimism accompanying the invention of software, nuclear power, and space travel—there’s just not as much scientific frontier remaining. Moreover, is it really plausible that many of these re-directed liberal arts students would be capable enough to substantially alter the boundaries of civilization’s technological capabilities? Most STEM PhDs don’t even enjoy such a fate, and demand for this high-end work has been overstated. Already twenty-five years ago physics professor Jonathan Katz was famously cautioned students not to become a scientist because of the lackluster job prospects (due to a glut of scientists) and shortage of productive research opportunities: he warned that they’d languish in an endless series of postdocs, hoping for some slim chance at permanent employment, until ultimately they might just defect and become software developers.

What can be done? The first thing for any young person (which means anyone who does not have a permanent job in science) to do is to pursue another career. . . . I have known more people whose lives have been ruined by getting a Ph.D. in physics than by drugs.

The glut of scientists doesn’t square with the upbeat narrative around STEM. Even if academia is over-saturated, shouldn’t there be overwhelming opportunities for these folks in industry as scientists rather than as programmers, given how supposedly useful their scientific training is? Some of these people do land remunerative careers in finance or software, but doesn’t that misalignment between their expertise and their eventual job just showcase how even STEM education is more about signaling or ability rather than skill acquisition? And is some physics washout deciphering profitable ways to execute stock trades milliseconds faster than other people really the material progress people are imagining that STEM education generates?

The Downsides to STEM Boosterism

I’ve mainly focused on outlining how STEM underdelivers on its supposed societal benefits, but promoting STEM also has downsides, including the talent mismatching mentioned above. We’ve already been down this road before by encouraging everyone to go to college regardless of their aptitude. That was a mistake. As

Arnold Kling

describes, colleges are now brimming with folks who shouldn’t be there:

Fundamentally, there are too many people on a college campus who don’t belong there.

When I was an adjunct at George Mason, most of my students could not write or do math. Reading their essays or grading their exams was painful. I wanted to forward them to the admissions department and ask, “What are you doing?” It was the rare student who could actually think at a level that justified being in a college-level course.

Similarly, many people who don’t currently study STEM would struggle in STEM, and cramming people lacking the requisite ability into these programs is unwise. Plus, STEM promotion further enshrines the narrow-mindedness about human value and societal status that Freddie deBoer writes about. And, as I’ve pointed out before, the elimination of liberal arts seems kind of dystopic anyway:

[A]re the rest of us certain that we need to live in such an unforgiving and cutthroat late-capitalist nightmare wherein every person needs to attempt majoring in something like computer science or accounting, regardless of their talents or interests, and where not even college can shelter fleeting moments of idealism? With all the advances in society, we somehow can’t permit the lassitude and curiosity we afforded 40 years ago? Maybe that’s pushing us in an economically efficient direction, but is it pushing us in a positive direction?

The primary drawback to evangelizing STEM, however, besides the ensuing havoc of over-stuffing those programs with less talented/enthusiastic students, the opportunity cost of forgoing more intellectually enriching disciplines, and the further winnowing of societal status to a minuscule fraction of the population, is mostly that STEM is just so toilsome and strenuous: the Potemkin-village-like routine of pretending to traffic in practical abilities wouldn’t be as galling if it weren’t so damn intense. We should be clever enough to avoid such hellish squandering of our cognitive resources.

Conclusion

My position on STEM isn’t my only opinion about education that’s changed—the importance of signaling and innate talent have upped my cynicism about education overall. For example, I’ve also lost confidence in the potential of educational reforms like school vouchers, and this essay is more about countering misplaced positivity around STEM than about praising the humanities. We can still improve education, but there's too much of it, and it’s swallowing the culture and eating our youth.

Also, the humanities have become a scapegoat, and the blinkered optimism about STEM has become a way to avoid reckoning with the challenges surrounding the variability and inherent limitations of the human mind. STEM boosterism indiscriminately encourages ill-fitting talent to funnel into challenging disciplines. Ultimately, the hunger to promote STEM education is predicated on a faulty presumption that the observed success of graduates is entirely due to teachable skills that are widely exportable to untapped populations. But really it’s mainly due to signaling and aptitude. And in examining the twin aims of university education—personal and intellectual enrichment—STEM is overestimated as a pathway towards the former and is typically a poor strategy for securing the latter.