The STEM curriculum itself could be seriously improved. Right now it's designed for the needs of a mechanical engineer circa 1955. Spends about 3x more time than necessary on elaborate trigonometry and calculating the volumes of increasingly absurd objects. Yet even the advanced students who take Calc BC won't come out knowing what a Lagrange multiplier is.
It's hard in ways that are sort of useless and neglects important concepts which then have to be learned in college.
Interestingly some of the math education reform efforts that inspire controversy (because they get rid of middle school tracking) do actually make some good moves in this direction, incorporating probability, statistics, and computation earlier.
I'd also add that we should introduce statistical methods way earlier. Quality control is all about understanding the statistical range of outcomes for many sub-components or processes, and how those combine together into your final rate of success or failure.
Even better if applied statistics is required in advanced physics, chemistry, geography etc. Studying statistical methods in isolation makes statistics hard to love.
Totally -- teach stats in the context of how it's actually used! But that's true of basically ALL math. You can teach geometry, algebra, and trig, all in the context of how you use those disciplines to design a bridge that won't fall down. Cantilevers and arches! Very few people care about the elegance of the math by itself. LOTS of people love thinking about building cool stuff -- witness the enduring popularity of David Macaulay.
Interesting parallels to law school there - spends about 3x more time than necessary on elaborate, increasingly absurd fine points of caselaw, but very little time talking about practical skills or even everyday, real-world concepts. That's gotten a bit better in some places with more emphasis on clinical education and experience from summer jobs, but antiquated curriculum might be a problem across disciplines.
I actually think there is too much practicality in the STEM curriculum — lots of drilling hand-computations and solving real world problems, but ones that that just aren’t very relevant any more.
It’s like if law schools devoted time to teaching how to manage scriveners and use card catalogs.
The purpose of modern engineering education is to get people to the point that they can tell if the computer is giving them the wrong answer. ;)
(Because you made a typo, or there's a bug in the software your co-worker wrote, or the model depends on assumptions that don't hold in your particular case, or any number of other ways Murphy's Law can strike and make the computer's answer ridiculous.)
Another thing that is happening in the controversial Virginia reform package, as I understand it, is the elimination of discrete years for Algebra, Geometry, Pre-Calc, etc., and instead a broad-based ladder of complexity throughout the high school years that incorporates all different kinds of math. Which makes way more sense to me, frankly. Freshmen should learn the easy bits of everything, and sophomores learn the somewhat harder bits, and then juniors master it, and then seniors take calculus or something.
Calculus has its easy parts too. But yes, the current scheme makes only slightly more sense than learning French by studying nouns one year and verbs the next.
This fact, and its absence in the essay, makes me really skeptical of the rest of your argument.
> But the fact that such a high percentage of our grad students in STEM fields are from overseas suggests that there’s huge room for improvement in how well we train our own workers.
Nope, it's because for most people, grad school is a shit choice. You spend several years making ~10k a year, maybe 20k tops, in what is really a low-paid research position. This is a good deal if you're looking to get a visa, but otherwise its not that great of a gig. That's why the positions are so heavily staffed by foreigners: they're getting a visa out of the deal.
I don't think there's an real evidence of a stem shortage of American workers. The more accurate story is that there just aren't a ton of STEM jobs available for most Americans, unless they are willing to relocate to a relatively small number of really expensive cities. If you want more Americans working in STEM, i think you need more STEM jobs distributed around the country, and you probably want them to pay more as well.
Of course there's also the question of the terrible mismatch between the incentives of the higher education system, and those of the students going in. The university system touts itself as being the magic path out of poverty, while giving many students little valuable in exchange for heavy amounts of debt.
All these facts - the prevalence of people getting stem degrees and then not working in stem, grad school being a bad deal for the most part - the broken incentives of universities - they're all pretty obvious to anyone willing to look for them. But since these political coalitions tend to line up on the 'blue' side of the aisle, most journalists aren't writing about them, and the narrative manufacturing machines are ignoring jsut how awful the american education system is failing people.
If you want to improve STEM AND fight inequity, we should retribute wealth away from stanford, MIT, harvard, and yale, and start building nuclear power plants in every state. We also need _some_ system that aligns incentives between 'people who are young and want to be trained in order to get good paying jobs' and 'people who can do the training' - because the university system sure as hell isn't it. Once engineering jobs look high status, and young people are getting advice from people with skin in the game - people will start to migrate towards the good paying enginerring jobs.
Sure those stem graduates aren’t going into “stem jobs”, But what they are going into his technical management, and other high paying careers. The survey is more a representation of what should be defined as a stem job than it is a treaties on stem graduates.
They're not working at Starbucks like your history major’s.
Is there some evidence of a supply crunch that i'm missing?
I get the impression that there's a never demand for software engineers, mostly concentrated in a few small cities. As a software engineer, i find this pretty swell. But the never-ending demand seems to be largely limited to software engineers.
But is there really some huge demand crunch for, say, mechanical engineers or electrical engineers?
It's not like there are a bunch of fabs in America that can't find engineers to run them. If anything, the problems with American science education seem to be that being a scientist isn't a high paying, high status career, compared to being, say, a lawyer or a doctor.
STEM grad students make more than 10 -20 k, though? I got six or seven offers from math PhD programs, ranging from 17k to 34k, most somewhere 20-25k.
I took the 34k one, and with a relatively low cost of living and the ability to defer payments/not collect interest on my federal student loans, it really doesn't seem like that horrible a financial choice. If I get a research job afterwards, cool. If not, I just delayed getting a programming job or something for five years or so, and got to do some more interesting work in the meantime.
When I was a grad student at UC Berkeley in 1999-2000, I was making $18k, between a stipend and Research Assistant pay. I dropped out to take a job in industry after barely more than a semester. YOU try living in Berkeley at $18k a year.
Yeah, that's not great! I think my offer from UC Davis was towards the lower end of what I was offered (maybe ~22k), and the UC system pays grad students shit in general, especially relative to cost of living in most of CA (hence the massive strikes last year, which imo were very justified).
I wasn't trying to say that grad student pay shouldn't be higher, just that the range is more like 17-30k rather than 10-20k for stem PhD programs.
The "STEM shortage" really tracks closer to a "TE shortage." Science and math programs are incredibly helpful in getting jobs in these fields, but are not as singularly focused on getting you a job in them compared to tech and engineering programs.
I accept the main premise of your essay, but. I'm an old guy; the Sputnik moment happened before high school, and my math instruction was changed accordingly. For me, it was great, once we got past the lunacy of set theory for 8th graders.
I agree that we need to make STEM more broadly attractive, while improving attitudes and educational persistence. But I think you disregard that social science research is characteristically politicized, tending to favor the politically-favored outcome (in this case, that tracking doesn't work). Indeed it's extremely important to improve the education of ALL kids, and to improve the education of the (say) 5% of kids who are likely to enter STEM vocations, but it''s also important to improve the education of the 0.1% of kids who are best positioned to make the most significant contributions to the advancing edge of scientific research. Studies of average performance of a large-ish population are unable to detect effects on a small subgroup.
I know plenty of people in my field of scientific research who have made major contributions without remarkable intellectual chops, because of some combination of tenacity, grit, and imagination. But I also know plenty of the "other kind." If you look through the history of science, you will be struck by the sizable fraction of scientists and mathematicians who were prodigies at the age of ten. We see nothing weirdly unwholesome about musicians like Mozart. Why is this notion toxic when applied to STEM?
How much did being pushed into accelerated classes help those prodigies, though? I know some of those folks, and they needed no discovering or pushing.
It's not that easy to answer. Studies of mean tendencies don't properly address the issue, and IMHO social scientists tend to be strongly biased toward the null for political reasons, leading me to distrust their conclusions. I was not exceptionally productive in my professional life, for many reasons. One contributor (though not the main one) was that public school was so easy for me that I was unprepared for top-level college classes. I'd call most of that idiosyncratic.
What I can say based on my own experience that's somewhat generalizable is that it was painfully boring to be in a math class where the teacher (appropriately) had to spend her time explaining multiplication when I was already far beyond that, and I shall never forget the thrill I felt when (as part of the Sputnik hysteria) we got tracked into a special math class with "SMSG" (School Mathematics Study Group) texts prepared by mathematicians and designed for kids but at a serious level. Part of the success certainly was attributable to one exceptionally talented teacher, who would have been effective no matter what she taught.
I agree with you that enhancing STEM instruction and expectations for a large group of students (all? 50%? 25%?) is important, though surely (in your example) the security guards didn't need to be STEM-trained. That leaves the issue of the degree to which additional attention to the very top of the class might provide additional benefits, as hard to know with certainty, but it seems common sense that one might reasonably invest some extra resources into the education of the top (say) 1% of kids.
This is why the "care work is infrastructure" messaging makes me uneasy. I'm completely supportive of getting funding for childcare, elder care, the whole liberal wishlist by any means necessary, and I'm also in favor of thinking of infrastructure as more than roads. But even pre-fabs, someone has to keep the lights on. Someone has to filter the drinking water and maintain the sewers and update the building codes. It's incredibly easy to take the basic physical elements of civilization for granted, and we need people in every community in the country who are going to be ready to step up and learn how to run and maintain those systems (most of which, of course, are now inextricably computerized as well). Respect for this as unique and vital work is, I think, hugely important.
One argument that I see particularly in computer science/software engineering is the importance of humanities education and the few (but not zero) number of English/Poli Sci majors who take a coding bootcamp and land a tech job.
I took a few general courses in addition to my engineering major which I enjoyed. I wonder if you could speak more to the types of education curriculums that could better entice people to a STEM degree even as we preserve many of the teachings of humanities courses.
Philosophy-turned CS/math double major here. It is unbelievably difficult to get hired as a humanities major with just a coding bootcamp. Tech employers are incredibly risk-averse when it comes to entry-level positions.
Using programming as an example, if you hire a CS grad from a top 20 school then it doesn't matter how it turns out. You did the "reasonable" thing and if they don't do well no one is going to blame you. But if you hire a sociology major who has been coding for just a couple years, it is 100% on you if they struggle.
I think we could entice a lot more people to take STEM classes if the bar for entry-level jobs wasn't so absurdly high. Maybe the government could fix this with an increased research budget? Research generates grunt work that researchers could delegate to entry-level positions. Some of the people in those entry-level jobs will do great and go on to have wonderful STEM careers, and in the meantime we free up researchers to be more productive.
When I was a young hotshot, one of the most motivating things was finding out that there *was* so much more to be known. In other words, those accelerated classes, but also something most US high schools seem to avoid like the plague, namely early specialisation. I specialised in STEM at age 14 and 6 months (and no, this was not in some totalitarian pressure cooker; I also took German, just for fun). Seven years later, I moved to the USA.
Don't forget that tests on specific subjects will, in reality, measure motivation blended with ability.
To add to this, those accelerated classes had a social benefit of showing kids that didn't come from academic or professional families what other opportunities are out there. When your parents work in a factory or as a hot tub salesperson, you really don't know much about jobs that require high educational attainment (outside of healthcare and teaching). Making friends in these classes becomes an introduction to the professional world in many ways.
- "In recent years there’s a push to de-emphasize STEM, mostly out of egalitarian concerns." I get that pursuing STEM is unfashionable or even sometimes problematic in some circles, but at large I haven't seen that same trend. In fact, compared to when I was in college (2010ish) it feels like it is more popular to pursue STEM now than ever before: https://www.nap.edu/read/25038/chapter/4#43
Great article! You hit the nail right on the head! My major was math achieved in 1974. Lead to an interest in programming using assembly language on a PDP-8 given to the Chemistry department by DEC. When IBM launched their PC I was one of the few at work that used it.I used a FORTRAN compiler to produce code that made impact testing more accurate.
I agree with the problem but disagree with the solution. STEM "talent" probably more has more to do with interest and confidence than innate ability, and it's still worth identifying and encouraging it early in life. A solid foundation in STEM for all students is also a good idea, but I doubt there would value in trying to herd a big portion of the population into specializing in it. Let the kids who feel they're good at it move ahead.
The NYC decision re G&T admissions was only for this year, because they felt they couldn't safely organize the testing. There's continuing interest in introducing additional criteria besides the test, but for now the plan is to switch back next schoolyear to the test as the only admissions criteria. There's a brand new but possibly short term DOE chief who has talked about her interest in changing G&T admissions, but probably nothing will be decided until the new mayor comes in late this year, who could put in another new DOE chief. My daughter's in G&T and at least in NYC the stereotype that it's all white doesn't hold true. There are also a lot of accelerated charter schools in NYC, which do admissions by lottery, but the mere fact that to enter the lottery parents have to want that kind of school for their kid means these schools also tend to attract "talented" kids.
Just to clarify I certainly agree with the thrust of the op-ed that as a society we need to bust the myth of the born math whiz. Really anyone can do it. But I doubt we really need big numbers of specialists. And we need to bust it also for the math whizzes. I was one and it took me till my second year of college I didn't actually want to do that with my life.
NYC might be a good place to study what works and what doesn't, since they do a lot of different things here. There's an unusual amount of variety and specialization in the public system and a large diverse set of private schools. I was talking about accelerated learning generally - sorting kids especially good at and interested in STEM doesn't really begin till middle school, at least in the public system. There's a G&T K-8 school that calls itself STEM but the admission test is standard, half reading comprehension and half IQ-test-like pattern comprehension.
The theme of Scott Rozelle's "Invisible China" is that the economy is at risk because rural children, who comprise 70% of children in China, are not educated properly even to high school level. There are classrooms - but what goes on inside them is without value. Rozelle fears that this will lock China in the middle income trap. This is a version of Noah's piece. You have to get good STEM teachers, which may not be easy.
Chinese students are learning math while American students are learning to decolonize their privilege. America had a choice between equity and merit and we chose poorly.
Your daughter's high school texts are what, 20-30 years later than your college texts? We have come a long way generally, which is good. I think the general concern that we aren't prioritizing learning outcomes is fair - the USA isn't keeping up on outcomes
Why do we take calculus in high school? I do have a STEM job and don't think being able to do univariate calculus has ever helped me. It would've been better if we made everyone take AP statistics and I could've taken a discrete logic or real analysis class.
Because most engineering before the 1990s required a large number of people capable of crunching numbers on physical phenomenon acting on physical structures. Since then, engineering has started to drift more towards data organization and analysis. This is mostly due to a vast increase in computing power available to every high-tech worker.
Now the questions are less like, "What is the answer to build a widget within these constraints?" Instead it's, "Which of these several widget designs is best for our application?"
STEM background here (Ph.D. physical chemistry), but I became a Wall Street banking lawyer. A number of my colleagues were also STEM folk. Why? Better pay; easier work; less competition. The only rational people who end up in STEM are those with an extremely powerful taste for it, and those with no alternatives (immigrants, folk with poor social capital, etc.)
Hi Noah, I work in a fab (which of course does not automatically make me authoritative, but it does give me a certain perspective). I largely agree with your recommendations on education reform. However I'd like to make some comments on the industrial side.
1) There is not only a need for PhDs in the fab, but also shift personnel, including folks who are good at fixing things with their hands, or who simply have brute strength and can do manual labor. In fact these folks can be harder to find, since for example auto mechanics usually don't think of themselves as well qualified for fab work. I believe Varian Semiconductor (before they were purchased by the Evil Empire a.k.a. Applied Materials) ran an ad campaign which more or less said, "if you can fix a car you can fix an ion implanter." In my experience working on fab equipment is even easier than working on cars, since it is designed to be routinely maintained.
2) I think much of the movement of semiconductor manufacturing out of the US and into Asia has more to do with macroeconomics in general than STEM education in particular, although I'm sure you'd understand that better than me 😊. It is astonishingly cheaper to run a leading edge fab (or any fab really) where labor is so much cheaper, and so much of the rest of the supply chain is colocated. I do struggle to source parts and chemicals from within the US... It is frequently so much cheaper to buy them overseas, and I have to go to lengths to justify the added expense of buying local to my superiors.
3) Fabs produce truly horrible waste that will destroy the surrounding environment if not properly contained and/or passivated. It is always cheaper to vent your exhaust directly to the atmosphere or house exhaust than to install abatement units, for example. In the US companies go to great length to understand the unique waste properties of every process and dispose accordingly, as it should be. Many of our foreign competitors have no regard for these practices, which gives them a competitive advantage in a marketplace that does not price in externalities.
Instead of identifying the best students and accelerating their education, why not identify the best students and set them to tutoring those who are less good? It can raise the average standard, teaching will improve the abilities of the gifted ("if you want to learn, teach") and it will promote solidarity.
Agree. It might do more towards manning the fabs though. I feel like we need to see education as more of a team sport to achieve that goal. At least in America or Europe.
G&T programs generally start much younger. By the time kids get old enough to teach, they can usually self-study.
In my opinion the anti-G&T people aren't thinking straight. They want to even the gap between the average test scores of different racial groups, and they see disproportionate numbers of white and Asian kids in G&T, so they think G&T is a way those races are rigging the system to give themselves an unfair advantage. But the problem is not G&T: it's the way admissions are run; the locations of G&T programs and difficulties for low-income people to get their kids to them; the cultural stereotypes that make less-educated, lower-income parents less likely to think their kids should apply. There is always going to be a very big influence of parents on kids' education performance and because of that some extent of intergenerational low-income trap. But public G&T programs are one of the best escape routes. School districts that shut them down will be reducing the numbers of their low-income kids who go on to nationally top-ranked universities, don't doubt it. And it's not to going to help the kids who are struggling at all.
As a Canadian high schooler I routinely tutored other students directed to me by teachers. This causes stigma on both sides as I'm identified as the goodie 2 shoes, teachers pet, and the other student is either embarrassed to be taught by someone 2 years younger than them, or hostile because they didn't want to be there. A partner system in class works better, as well as anonymous outsider / adult.
My criticism of the Virginia program is a bit different than those that you mentioned. It requires one to rethink the "gifted" nomenclature. Simply put, of all the smart people I've interacted with, I can count those with unexplained (academic) talent using only the number zero. Instead, everyone filters into roughly 3 categories: high socioeconomic status (aka SES), high motivation + mentor, and clear cases of neurodivergence (these people normally have severe developmental issues outside of their "gift(s)"). I'm sure there are rare cases of unmotivated, poor, yet level-headed kids out there, but we're talking about 1 in 100 million or 1 billion, not 30 of them taking the same AP physics class in a Cleveland, Ohio magnet school.
If one uses that model, the Virginia decision looks a bit different. High SES kids don't suffer at all. They'll always get resources to succeed. Motivated and mentored kids will suffer some, since excess motivation might become excess boredom in many areas, but mentors may find even more motivated kids who were otherwise discounted too early in the old system. This will hurt and help different individuals, but will probably be awash as a whole.
For the neurodivergent, though, they will lose out on building on their strengths early without additional resources being devoted to their weaknesses. The US mental health system just isn't prepared to help most of them. They'll end up bored and bullied (by kids and *teachers*), along with given fewer opportunities to understand where their strengths can go. I fear by the time they get to 11th grade, many will have decided to give up on further (math) education to chase other interests that have programs that are more eager to foster their strengths. Subjects where exceptional performance gets them more positive attention.
If the goal is to get even more STEM professionals, chasing neurodivergent kids into sports and fine arts will definitely not help towards that goal.
I don't know what's going on in Virginia, only in New York. Here while the talk in the public sphere is about changing the admission criteria for G&T from testing-alone, which could be good, the rumor among G&T teachers is that higher-ups want to kill the program. Which would be a big shame. I think there are kids who should be put into accelerated learning, because they will be bored and unmotivated in regular classes. It's much more to do with parents' education levels and personalities than economic status - poor artists with PhDs parents typically have accelerated-learning kids. And then there are "twice exceptional" kids who benefit greatly from schools that recognize their combination of special talents and difficulties.
The NYC public G&T system is not equipped to handle "twice exceptional" kids. There are good private schools doing it, and they give scholarships and there are some public subsidy programs. No one here is trying to get rid of those. But parents have to be very aware and proactive to take advantage of them, so a lot of kids who should be in "twice exceptional" programs are sadly not.
The G&T system is under attack with the idea that it draws resources out of general education to kids who are considered to need it less. But in my view the G&T's inequities are the same as the overall system. Wealth being concentrated in lower Manhattan most resources have traditionally gone there. The best G&T schools are there. And with its density of people and subway trains, G&T schools are closer together and easier to reach. In upper Manhattan and the outer boroughs, subway lines are usually no help because they run in radials to lower Manhattan and your closest G&T school will probably not be on your line. So people without cars simply can't get their kids to G&T schools.
There's also the strong tradition here of discipline-oriented, accelerated-learning charter schools. These are especially popular with blacks and latinos, who often prefer them over regular public G&T. Even so in NYC outer boroughs G&T classes run from mostly to nearly all non-white, but the kids nevertheless typically have parents with higher education.
I'd very much like to see these programs improved not demolished, which in my opinion would be a terrible shame.
The STEM curriculum itself could be seriously improved. Right now it's designed for the needs of a mechanical engineer circa 1955. Spends about 3x more time than necessary on elaborate trigonometry and calculating the volumes of increasingly absurd objects. Yet even the advanced students who take Calc BC won't come out knowing what a Lagrange multiplier is.
It's hard in ways that are sort of useless and neglects important concepts which then have to be learned in college.
Interestingly some of the math education reform efforts that inspire controversy (because they get rid of middle school tracking) do actually make some good moves in this direction, incorporating probability, statistics, and computation earlier.
I'd also add that we should introduce statistical methods way earlier. Quality control is all about understanding the statistical range of outcomes for many sub-components or processes, and how those combine together into your final rate of success or failure.
I truly believe statistics should be a high school graduation requirement.
Even better if applied statistics is required in advanced physics, chemistry, geography etc. Studying statistical methods in isolation makes statistics hard to love.
Totally -- teach stats in the context of how it's actually used! But that's true of basically ALL math. You can teach geometry, algebra, and trig, all in the context of how you use those disciplines to design a bridge that won't fall down. Cantilevers and arches! Very few people care about the elegance of the math by itself. LOTS of people love thinking about building cool stuff -- witness the enduring popularity of David Macaulay.
Yeah.
Interesting parallels to law school there - spends about 3x more time than necessary on elaborate, increasingly absurd fine points of caselaw, but very little time talking about practical skills or even everyday, real-world concepts. That's gotten a bit better in some places with more emphasis on clinical education and experience from summer jobs, but antiquated curriculum might be a problem across disciplines.
I actually think there is too much practicality in the STEM curriculum — lots of drilling hand-computations and solving real world problems, but ones that that just aren’t very relevant any more.
It’s like if law schools devoted time to teaching how to manage scriveners and use card catalogs.
The purpose of modern engineering education is to get people to the point that they can tell if the computer is giving them the wrong answer. ;)
(Because you made a typo, or there's a bug in the software your co-worker wrote, or the model depends on assumptions that don't hold in your particular case, or any number of other ways Murphy's Law can strike and make the computer's answer ridiculous.)
Another thing that is happening in the controversial Virginia reform package, as I understand it, is the elimination of discrete years for Algebra, Geometry, Pre-Calc, etc., and instead a broad-based ladder of complexity throughout the high school years that incorporates all different kinds of math. Which makes way more sense to me, frankly. Freshmen should learn the easy bits of everything, and sophomores learn the somewhat harder bits, and then juniors master it, and then seniors take calculus or something.
Calculus has its easy parts too. But yes, the current scheme makes only slightly more sense than learning French by studying nouns one year and verbs the next.
Most US STEM graduates aren't working STEM jobs:
https://www.census.gov/newsroom/press-releases/2014/cb14-130.html
This fact, and its absence in the essay, makes me really skeptical of the rest of your argument.
> But the fact that such a high percentage of our grad students in STEM fields are from overseas suggests that there’s huge room for improvement in how well we train our own workers.
Nope, it's because for most people, grad school is a shit choice. You spend several years making ~10k a year, maybe 20k tops, in what is really a low-paid research position. This is a good deal if you're looking to get a visa, but otherwise its not that great of a gig. That's why the positions are so heavily staffed by foreigners: they're getting a visa out of the deal.
I don't think there's an real evidence of a stem shortage of American workers. The more accurate story is that there just aren't a ton of STEM jobs available for most Americans, unless they are willing to relocate to a relatively small number of really expensive cities. If you want more Americans working in STEM, i think you need more STEM jobs distributed around the country, and you probably want them to pay more as well.
Of course there's also the question of the terrible mismatch between the incentives of the higher education system, and those of the students going in. The university system touts itself as being the magic path out of poverty, while giving many students little valuable in exchange for heavy amounts of debt.
All these facts - the prevalence of people getting stem degrees and then not working in stem, grad school being a bad deal for the most part - the broken incentives of universities - they're all pretty obvious to anyone willing to look for them. But since these political coalitions tend to line up on the 'blue' side of the aisle, most journalists aren't writing about them, and the narrative manufacturing machines are ignoring jsut how awful the american education system is failing people.
If you want to improve STEM AND fight inequity, we should retribute wealth away from stanford, MIT, harvard, and yale, and start building nuclear power plants in every state. We also need _some_ system that aligns incentives between 'people who are young and want to be trained in order to get good paying jobs' and 'people who can do the training' - because the university system sure as hell isn't it. Once engineering jobs look high status, and young people are getting advice from people with skin in the game - people will start to migrate towards the good paying enginerring jobs.
Sure those stem graduates aren’t going into “stem jobs”, But what they are going into his technical management, and other high paying careers. The survey is more a representation of what should be defined as a stem job than it is a treaties on stem graduates.
They're not working at Starbucks like your history major’s.
My direct boss at my STEM job has a BA in history...
My brothers an engineer for Boeing. He jokes about how management makes more than the engineers themselves.
Which is also why Boeing has a lot of drama lately.
We need both supply and demand, sure. No argument there. Not sure about the nuclear power plant thing though, I don't think that's the ticket here.
Is there some evidence of a supply crunch that i'm missing?
I get the impression that there's a never demand for software engineers, mostly concentrated in a few small cities. As a software engineer, i find this pretty swell. But the never-ending demand seems to be largely limited to software engineers.
But is there really some huge demand crunch for, say, mechanical engineers or electrical engineers?
It's not like there are a bunch of fabs in America that can't find engineers to run them. If anything, the problems with American science education seem to be that being a scientist isn't a high paying, high status career, compared to being, say, a lawyer or a doctor.
STEM grad students make more than 10 -20 k, though? I got six or seven offers from math PhD programs, ranging from 17k to 34k, most somewhere 20-25k.
I took the 34k one, and with a relatively low cost of living and the ability to defer payments/not collect interest on my federal student loans, it really doesn't seem like that horrible a financial choice. If I get a research job afterwards, cool. If not, I just delayed getting a programming job or something for five years or so, and got to do some more interesting work in the meantime.
When I was a grad student at UC Berkeley in 1999-2000, I was making $18k, between a stipend and Research Assistant pay. I dropped out to take a job in industry after barely more than a semester. YOU try living in Berkeley at $18k a year.
Yeah, that's not great! I think my offer from UC Davis was towards the lower end of what I was offered (maybe ~22k), and the UC system pays grad students shit in general, especially relative to cost of living in most of CA (hence the massive strikes last year, which imo were very justified).
I wasn't trying to say that grad student pay shouldn't be higher, just that the range is more like 17-30k rather than 10-20k for stem PhD programs.
My numbers are probably showing my age a bit - i started grad school almost 15 years ago.
The "STEM shortage" really tracks closer to a "TE shortage." Science and math programs are incredibly helpful in getting jobs in these fields, but are not as singularly focused on getting you a job in them compared to tech and engineering programs.
I accept the main premise of your essay, but. I'm an old guy; the Sputnik moment happened before high school, and my math instruction was changed accordingly. For me, it was great, once we got past the lunacy of set theory for 8th graders.
I agree that we need to make STEM more broadly attractive, while improving attitudes and educational persistence. But I think you disregard that social science research is characteristically politicized, tending to favor the politically-favored outcome (in this case, that tracking doesn't work). Indeed it's extremely important to improve the education of ALL kids, and to improve the education of the (say) 5% of kids who are likely to enter STEM vocations, but it''s also important to improve the education of the 0.1% of kids who are best positioned to make the most significant contributions to the advancing edge of scientific research. Studies of average performance of a large-ish population are unable to detect effects on a small subgroup.
I know plenty of people in my field of scientific research who have made major contributions without remarkable intellectual chops, because of some combination of tenacity, grit, and imagination. But I also know plenty of the "other kind." If you look through the history of science, you will be struck by the sizable fraction of scientists and mathematicians who were prodigies at the age of ten. We see nothing weirdly unwholesome about musicians like Mozart. Why is this notion toxic when applied to STEM?
How much did being pushed into accelerated classes help those prodigies, though? I know some of those folks, and they needed no discovering or pushing.
It's not that easy to answer. Studies of mean tendencies don't properly address the issue, and IMHO social scientists tend to be strongly biased toward the null for political reasons, leading me to distrust their conclusions. I was not exceptionally productive in my professional life, for many reasons. One contributor (though not the main one) was that public school was so easy for me that I was unprepared for top-level college classes. I'd call most of that idiosyncratic.
What I can say based on my own experience that's somewhat generalizable is that it was painfully boring to be in a math class where the teacher (appropriately) had to spend her time explaining multiplication when I was already far beyond that, and I shall never forget the thrill I felt when (as part of the Sputnik hysteria) we got tracked into a special math class with "SMSG" (School Mathematics Study Group) texts prepared by mathematicians and designed for kids but at a serious level. Part of the success certainly was attributable to one exceptionally talented teacher, who would have been effective no matter what she taught.
I agree with you that enhancing STEM instruction and expectations for a large group of students (all? 50%? 25%?) is important, though surely (in your example) the security guards didn't need to be STEM-trained. That leaves the issue of the degree to which additional attention to the very top of the class might provide additional benefits, as hard to know with certainty, but it seems common sense that one might reasonably invest some extra resources into the education of the top (say) 1% of kids.
This is why the "care work is infrastructure" messaging makes me uneasy. I'm completely supportive of getting funding for childcare, elder care, the whole liberal wishlist by any means necessary, and I'm also in favor of thinking of infrastructure as more than roads. But even pre-fabs, someone has to keep the lights on. Someone has to filter the drinking water and maintain the sewers and update the building codes. It's incredibly easy to take the basic physical elements of civilization for granted, and we need people in every community in the country who are going to be ready to step up and learn how to run and maintain those systems (most of which, of course, are now inextricably computerized as well). Respect for this as unique and vital work is, I think, hugely important.
Sure. That's part of the "two-track economy" plan, I think. Which has its issues, but is generally a smart idea.
One argument that I see particularly in computer science/software engineering is the importance of humanities education and the few (but not zero) number of English/Poli Sci majors who take a coding bootcamp and land a tech job.
I took a few general courses in addition to my engineering major which I enjoyed. I wonder if you could speak more to the types of education curriculums that could better entice people to a STEM degree even as we preserve many of the teachings of humanities courses.
Yeah that's very interesting stuff.
Philosophy-turned CS/math double major here. It is unbelievably difficult to get hired as a humanities major with just a coding bootcamp. Tech employers are incredibly risk-averse when it comes to entry-level positions.
Using programming as an example, if you hire a CS grad from a top 20 school then it doesn't matter how it turns out. You did the "reasonable" thing and if they don't do well no one is going to blame you. But if you hire a sociology major who has been coding for just a couple years, it is 100% on you if they struggle.
I think we could entice a lot more people to take STEM classes if the bar for entry-level jobs wasn't so absurdly high. Maybe the government could fix this with an increased research budget? Research generates grunt work that researchers could delegate to entry-level positions. Some of the people in those entry-level jobs will do great and go on to have wonderful STEM careers, and in the meantime we free up researchers to be more productive.
Perhaps it's survivorship bias, where I only see the successful boot campers at my company and not those who failed.
But I also do see existing opportunities for entry-level folks who can just build a website or an app without needing a deep engineering background.
When I was a young hotshot, one of the most motivating things was finding out that there *was* so much more to be known. In other words, those accelerated classes, but also something most US high schools seem to avoid like the plague, namely early specialisation. I specialised in STEM at age 14 and 6 months (and no, this was not in some totalitarian pressure cooker; I also took German, just for fun). Seven years later, I moved to the USA.
Don't forget that tests on specific subjects will, in reality, measure motivation blended with ability.
To add to this, those accelerated classes had a social benefit of showing kids that didn't come from academic or professional families what other opportunities are out there. When your parents work in a factory or as a hot tub salesperson, you really don't know much about jobs that require high educational attainment (outside of healthcare and teaching). Making friends in these classes becomes an introduction to the professional world in many ways.
You're one of the top quality blue-pill dealers.
Three Things:
- "In recent years there’s a push to de-emphasize STEM, mostly out of egalitarian concerns." I get that pursuing STEM is unfashionable or even sometimes problematic in some circles, but at large I haven't seen that same trend. In fact, compared to when I was in college (2010ish) it feels like it is more popular to pursue STEM now than ever before: https://www.nap.edu/read/25038/chapter/4#43
- Mathematician Steven Strogatz facilitates a conversation about math in America here: https://www.youtube.com/watch?v=6T9kV11YCBg. He also has a pretty good podcast.
- Really like the topic choice. Hope to see it in future post.
He's my friend. I should interview him!
Great article! You hit the nail right on the head! My major was math achieved in 1974. Lead to an interest in programming using assembly language on a PDP-8 given to the Chemistry department by DEC. When IBM launched their PC I was one of the few at work that used it.I used a FORTRAN compiler to produce code that made impact testing more accurate.
Keep hammering on this topic!
I agree with the problem but disagree with the solution. STEM "talent" probably more has more to do with interest and confidence than innate ability, and it's still worth identifying and encouraging it early in life. A solid foundation in STEM for all students is also a good idea, but I doubt there would value in trying to herd a big portion of the population into specializing in it. Let the kids who feel they're good at it move ahead.
The NYC decision re G&T admissions was only for this year, because they felt they couldn't safely organize the testing. There's continuing interest in introducing additional criteria besides the test, but for now the plan is to switch back next schoolyear to the test as the only admissions criteria. There's a brand new but possibly short term DOE chief who has talked about her interest in changing G&T admissions, but probably nothing will be decided until the new mayor comes in late this year, who could put in another new DOE chief. My daughter's in G&T and at least in NYC the stereotype that it's all white doesn't hold true. There are also a lot of accelerated charter schools in NYC, which do admissions by lottery, but the mere fact that to enter the lottery parents have to want that kind of school for their kid means these schools also tend to attract "talented" kids.
I don't think we do a good job of identifying interest and intrinsic motivation. That's what I'd like to see more of.
Just to clarify I certainly agree with the thrust of the op-ed that as a society we need to bust the myth of the born math whiz. Really anyone can do it. But I doubt we really need big numbers of specialists. And we need to bust it also for the math whizzes. I was one and it took me till my second year of college I didn't actually want to do that with my life.
sorry: second year of college to realize I didn't
NYC might be a good place to study what works and what doesn't, since they do a lot of different things here. There's an unusual amount of variety and specialization in the public system and a large diverse set of private schools. I was talking about accelerated learning generally - sorting kids especially good at and interested in STEM doesn't really begin till middle school, at least in the public system. There's a G&T K-8 school that calls itself STEM but the admission test is standard, half reading comprehension and half IQ-test-like pattern comprehension.
The theme of Scott Rozelle's "Invisible China" is that the economy is at risk because rural children, who comprise 70% of children in China, are not educated properly even to high school level. There are classrooms - but what goes on inside them is without value. Rozelle fears that this will lock China in the middle income trap. This is a version of Noah's piece. You have to get good STEM teachers, which may not be easy.
Yes, China is not as much of an invincible powerhouse as we've been led to believe. They're just really big.
Chinese students are learning math while American students are learning to decolonize their privilege. America had a choice between equity and merit and we chose poorly.
Your daughter's high school texts are what, 20-30 years later than your college texts? We have come a long way generally, which is good. I think the general concern that we aren't prioritizing learning outcomes is fair - the USA isn't keeping up on outcomes
Why do we take calculus in high school? I do have a STEM job and don't think being able to do univariate calculus has ever helped me. It would've been better if we made everyone take AP statistics and I could've taken a discrete logic or real analysis class.
Because most engineering before the 1990s required a large number of people capable of crunching numbers on physical phenomenon acting on physical structures. Since then, engineering has started to drift more towards data organization and analysis. This is mostly due to a vast increase in computing power available to every high-tech worker.
Now the questions are less like, "What is the answer to build a widget within these constraints?" Instead it's, "Which of these several widget designs is best for our application?"
Yep, and American math education should get out of the ‘50’s because the world has changed.
STEM background here (Ph.D. physical chemistry), but I became a Wall Street banking lawyer. A number of my colleagues were also STEM folk. Why? Better pay; easier work; less competition. The only rational people who end up in STEM are those with an extremely powerful taste for it, and those with no alternatives (immigrants, folk with poor social capital, etc.)
Yep, hence why we should tax the finance sector more.
Hi Noah, I work in a fab (which of course does not automatically make me authoritative, but it does give me a certain perspective). I largely agree with your recommendations on education reform. However I'd like to make some comments on the industrial side.
1) There is not only a need for PhDs in the fab, but also shift personnel, including folks who are good at fixing things with their hands, or who simply have brute strength and can do manual labor. In fact these folks can be harder to find, since for example auto mechanics usually don't think of themselves as well qualified for fab work. I believe Varian Semiconductor (before they were purchased by the Evil Empire a.k.a. Applied Materials) ran an ad campaign which more or less said, "if you can fix a car you can fix an ion implanter." In my experience working on fab equipment is even easier than working on cars, since it is designed to be routinely maintained.
2) I think much of the movement of semiconductor manufacturing out of the US and into Asia has more to do with macroeconomics in general than STEM education in particular, although I'm sure you'd understand that better than me 😊. It is astonishingly cheaper to run a leading edge fab (or any fab really) where labor is so much cheaper, and so much of the rest of the supply chain is colocated. I do struggle to source parts and chemicals from within the US... It is frequently so much cheaper to buy them overseas, and I have to go to lengths to justify the added expense of buying local to my superiors.
3) Fabs produce truly horrible waste that will destroy the surrounding environment if not properly contained and/or passivated. It is always cheaper to vent your exhaust directly to the atmosphere or house exhaust than to install abatement units, for example. In the US companies go to great length to understand the unique waste properties of every process and dispose accordingly, as it should be. Many of our foreign competitors have no regard for these practices, which gives them a competitive advantage in a marketplace that does not price in externalities.
I enjoy your writing, please continue 🙂
Instead of identifying the best students and accelerating their education, why not identify the best students and set them to tutoring those who are less good? It can raise the average standard, teaching will improve the abilities of the gifted ("if you want to learn, teach") and it will promote solidarity.
I did that all throughout school. It was useful. But it was only one piece of education; if that's all you ever do, you won't learn well.
Agree. It might do more towards manning the fabs though. I feel like we need to see education as more of a team sport to achieve that goal. At least in America or Europe.
G&T programs generally start much younger. By the time kids get old enough to teach, they can usually self-study.
In my opinion the anti-G&T people aren't thinking straight. They want to even the gap between the average test scores of different racial groups, and they see disproportionate numbers of white and Asian kids in G&T, so they think G&T is a way those races are rigging the system to give themselves an unfair advantage. But the problem is not G&T: it's the way admissions are run; the locations of G&T programs and difficulties for low-income people to get their kids to them; the cultural stereotypes that make less-educated, lower-income parents less likely to think their kids should apply. There is always going to be a very big influence of parents on kids' education performance and because of that some extent of intergenerational low-income trap. But public G&T programs are one of the best escape routes. School districts that shut them down will be reducing the numbers of their low-income kids who go on to nationally top-ranked universities, don't doubt it. And it's not to going to help the kids who are struggling at all.
As a Canadian high schooler I routinely tutored other students directed to me by teachers. This causes stigma on both sides as I'm identified as the goodie 2 shoes, teachers pet, and the other student is either embarrassed to be taught by someone 2 years younger than them, or hostile because they didn't want to be there. A partner system in class works better, as well as anonymous outsider / adult.
My criticism of the Virginia program is a bit different than those that you mentioned. It requires one to rethink the "gifted" nomenclature. Simply put, of all the smart people I've interacted with, I can count those with unexplained (academic) talent using only the number zero. Instead, everyone filters into roughly 3 categories: high socioeconomic status (aka SES), high motivation + mentor, and clear cases of neurodivergence (these people normally have severe developmental issues outside of their "gift(s)"). I'm sure there are rare cases of unmotivated, poor, yet level-headed kids out there, but we're talking about 1 in 100 million or 1 billion, not 30 of them taking the same AP physics class in a Cleveland, Ohio magnet school.
If one uses that model, the Virginia decision looks a bit different. High SES kids don't suffer at all. They'll always get resources to succeed. Motivated and mentored kids will suffer some, since excess motivation might become excess boredom in many areas, but mentors may find even more motivated kids who were otherwise discounted too early in the old system. This will hurt and help different individuals, but will probably be awash as a whole.
For the neurodivergent, though, they will lose out on building on their strengths early without additional resources being devoted to their weaknesses. The US mental health system just isn't prepared to help most of them. They'll end up bored and bullied (by kids and *teachers*), along with given fewer opportunities to understand where their strengths can go. I fear by the time they get to 11th grade, many will have decided to give up on further (math) education to chase other interests that have programs that are more eager to foster their strengths. Subjects where exceptional performance gets them more positive attention.
If the goal is to get even more STEM professionals, chasing neurodivergent kids into sports and fine arts will definitely not help towards that goal.
I don't know what's going on in Virginia, only in New York. Here while the talk in the public sphere is about changing the admission criteria for G&T from testing-alone, which could be good, the rumor among G&T teachers is that higher-ups want to kill the program. Which would be a big shame. I think there are kids who should be put into accelerated learning, because they will be bored and unmotivated in regular classes. It's much more to do with parents' education levels and personalities than economic status - poor artists with PhDs parents typically have accelerated-learning kids. And then there are "twice exceptional" kids who benefit greatly from schools that recognize their combination of special talents and difficulties.
The NYC public G&T system is not equipped to handle "twice exceptional" kids. There are good private schools doing it, and they give scholarships and there are some public subsidy programs. No one here is trying to get rid of those. But parents have to be very aware and proactive to take advantage of them, so a lot of kids who should be in "twice exceptional" programs are sadly not.
The G&T system is under attack with the idea that it draws resources out of general education to kids who are considered to need it less. But in my view the G&T's inequities are the same as the overall system. Wealth being concentrated in lower Manhattan most resources have traditionally gone there. The best G&T schools are there. And with its density of people and subway trains, G&T schools are closer together and easier to reach. In upper Manhattan and the outer boroughs, subway lines are usually no help because they run in radials to lower Manhattan and your closest G&T school will probably not be on your line. So people without cars simply can't get their kids to G&T schools.
There's also the strong tradition here of discipline-oriented, accelerated-learning charter schools. These are especially popular with blacks and latinos, who often prefer them over regular public G&T. Even so in NYC outer boroughs G&T classes run from mostly to nearly all non-white, but the kids nevertheless typically have parents with higher education.
I'd very much like to see these programs improved not demolished, which in my opinion would be a terrible shame.
Today in "the replies I get sometimes..."