Is there a shortage of STEM workers in the United States?

Via Matt, Hal Salzman, Daniel Kuehn, and Lindsay Lowell suggest there is no shortage because we do not observe significant wage rises for STEM workers.  That’s a fact worth knowing and I am happy to praise the study in that regard.  But it’s painting the basic worry into too narrow a box with its use of the word “shortage,” interpreted so literally.

The core claim is that STEM sectors will be those which produce the future social increasing returns for the economies which house them.  If true (I am not trying to prejudge this), that means we should invest in both more STEM workers and more complementary inputs, whether that be particle colliders, NIH funding, the right broadband infrastructure, legalizing driverless cars, better IP law, tougher schools, or whatever.  With the new, additional STEM workers, and the complementary inputs, America will (supposedly) be much better off.

To point out that the current supply of STEM workers stands in proper proportion to the other inputs suggests only that we are at a local optimum, not a global optimum.  Similarly, it could have been pointed out that, before the rise of Hyundai, South Korea had just the right number of auto workers (not many) for their factories (also not many).  That could have been true enough, but still investing in more auto factories and more auto workers was for Korea a very good path forward.

On top of all that, the report shows a worrying lack of concern about the notion of an economic margin.  Even without boosts in the complementary inputs, more STEM workers still can be put to good use, even if there is no “shortage” today.

Here is a related post by Adam Ozimek.  And Daniel Kuehn responds.


Or, growth and development is not about optimizing at the margin, but about shifting the cost/production possibility curve.

Is counter-cylical policy about marginal optimization or pushing the frontier outward? Is there reason to suspect that the two are related at all or that doing one implies doing less in the other?

Does this relationship change when interest rates are near lower bound or credit is constrained?

Interesting questions...

Quite simply: never reason from [the lack of] a price change.

Well that's what was reported, but we had a little more than that in there!

If we want to talk about other constraints on STEM in the U.S. I'm happy to have that conversation, but the mere existence of other constraints does not a shortage make.


As far as I can tell your study uses the fact that the wage premium for STEM grads hasn't grown as evidence that a shortage doesn't exist. But doesn't the presence of a large wage premium over other college grads imply a shortage? The market price of a STEM grad has remained persistently above cost since a STEM degree costs the same as a non-STEM degree.

If not a shortage, how else do you explain such a large wage premium? It may be the case that the magnitude of the shortage hasn't dramatically increased in the last 10-20 years but the persistant wage premium implies to me that we have been suffering from such a shortage for quite a long time.

No? Some activities are priced higher than others doesn't imply a shortage. The fact that caviar costs more than Saltines doesn't imply we have a caviar shortage. There is no reason to expect the equilibrium price point to be equal between different skill sets.

Caviar is more expensive than saltines because the production of a serving caviar requires substantially more real resources than the production of a serving of saltines.

The production of a STEM graduate requires roughly the same amount of real resources as the production of a liberal arts grad so absent some supply constraint the prices should be roughly equal.

The price of STEM labor is clearly not converging to the cost of producing STEM labor. How do you explain this?

But wouldn't we expect the return on STEM degrees to be the same as the return on other degrees? And if they have the same costs, then they must have the same price to create the same return.

To use your analogy, what if it costs $1 to produce both an ounce of caviar and an ounce of Saltines, but the Saltines are priced at $1.25 and the caviar is priced at $20. If that's the equilibrium, something must be constraining caviar production. Otherwise, people would just produce more caviar and make more money. Isn't that basically a shortage?

"The production of a STEM graduate requires roughly the same amount of real resources as the production of a liberal arts grad"

1. Take liberal arts student

2. Move into STEM program

3. Profit!

What could go wrong?

Degrees are not obsolete, they are useful, but they are definitely not required in IT field. I would rather have someone with a 180 IQ and a photographic memory over anyone who has a degree and a normal IQ. (For example) The idea that somehow we have to import hundreds of thousands of degrees is nonsense. In the ninties less than 20% of the people in the IT field even had degrees, and this has gone up to be about half, primarily because of H1B1s, not because they are required. Being really smart is what is nice to have, Degrees are optional.
People who work in programming over 20 years put in a lot more educational effort over those twenty years than it takes to get a four year degree.

We know that engineering graduates have had the highest starting salaries for some time, especially in the boring but lucrative petroleum field. We certainly accept that this is supply and demand. Is the argument that "sustained demand" is not "shortage?" I suspect a semantic argument is in play.

Thanks for linking! I agree with quite a bit of it. I've got thoughts on your post here:

The other context of all this is the immigration debate, and whether it makes sense to give high skill immigrants a fast track that other types of immigrants don't have. Is there cause for that? If you rely on labor shortages to answer "yes" it's probably worth reconsidering.

That does not mean, of course, that we can't also pursue what Ritwik suggests above. Growth is not about optimizing at the margin - but we weren't even talking about growth in this paper. I'd agree we need to shift the PPF.

The claim is that labor shortages don't provide an argument for giving one class of workers a leg up over other classes of workers.

I would also think that folks who care about income inequality would want to increase immigration of highly paid workers (not necessarily just STEM). That would lower salaries at the top and increase them at the bottom (higher wage workers would need services from lower wage natives who's supply was constrained).

I quite agree. The US should be making a major effort to increase immigration of highly skilled business executives and financiers, since the wages currently paid to such people indicate critical shortages of native-born talent. At the very minimum we should be bringing in several hundred thousand MBA graduates each and every year until the end of the decade. I can't understand why American politicians and economists are so unwilling to address this vital issue.

+1 mike shupp

Ya, I think this argument is really reaching. I get the argument, but I wouldn't want to defend it. It's more of a blunt instrument than the minimum wage, and THAT'S one blunt policy instrument.

Reminds me of an observation in either FIASCO or Liar's Poker about how there were tons of people willing to become i-banker or trader for a lot lot (lot) less than what the banks were paying... without the banks lowering their wage offers... One guy was trying to explain the bankers versus teachers salary via supply and demand and preferences and well either Frank Partnoy or Michael Lewis was having a field day making fun of that anonymous associate...

Actually, in a way this is happening. I hire financial/risk modellers and analysts, and an increasing % of applicants tend to be H1B.

Well you could start by importing a lot of doctors. They make up a good chunk of the top 1%.

I guess if you only care about the wages of the to .0001% yeah then it wold not have much impact

So everyone here agrees we have way too many low skilled workers and should stop importing them (wages too low, to much unemployment...etc). Just riffing off of the logic on offer.

This ignores global income inequality, and from that perspective low-skilled immigration would be favoured.

The argument for preferring one type of immigrant over another is that one type contributes more to the economy and global competitiveness than another. I'd rather have smart immigrants with masters degrees in STEM fields than uneducated (not meant pejoratively) laborers.

Disclaimer: I am pro-immigration for both types, but if I have to choose...

Matt Yglesias has been talking about your paper too and he is quite in favour of both type of immigration.

I am going to get around to writing a blog-post about immigration one of these days (basically, I am quite against for the EU or the USA at the present time). My reasoning is extremely blunt: There is no overall shortage since unemployment and underemployment is so high. Sector shortage are, of course, possible but I am indeed tempted to agree with you and your co-authors. Until and unless I see wage pressure in a given sector, I am not inclined to pre-suppose shortage or, actually, great benefits from immigration.

If an increase of labor supply was going to generate complementary benefits - how come present-day unemployed are not re-employed and generating them?

No, the only kind of "immigration" we ought to want right now are tourists and investors i.e. people willing to spend money.

If and when demand for labor manifests itself (or aging of the population creates an actual shortage), we can re-open our borders...

Fast tracking isn't really about bringing more immigrants, which are already coming anyway, but improve their position vs large companies.

Right now they can come as H1-Bs, and if they minimize their changes in employers, get permanent residency in 10-15 years, at which point they can compete evenly with Americans, instead of with depressed wages. Make the wait far shorter, and you'll see wages increase.

Tyler's argument was never about a labor shortage but about the externalities from more STEM.

The current US immigration system already tilts the playing field towards the unskilled through family reunification and through the habit of tending to regularize or to tolerate the illegals (mostly unskilled) who sneak in while making it hard for law-abiding skilled workers to get in the formal way. Reform and restrict family reunification and make deportation of illegals easier while expanding skilled visas and you get closer to a level playing field.

I'm glad Tyler mentioned tougher schools. There is already research showing that uneven grade inflation tends to hurt those majoring in STEM fields at the margin EVEN if on average STEM fields pay better. Economists understand this as the "lemons problem." Make schools tougher all around and diminish grade inflation and STEM will benefit more and as a bonus, the average college degree will become more meaningful.

re: "Tyler’s argument was never about a labor shortage but about the externalities from more STEM."

Right, but we think there are positive externalities from STEM - marginal social benefit is greater than marginal private benefit. So do you think the smartest way of going about fixing that is to work on the supply side or the demand side? I think the way to fix it is to address the demand side - we have ample evidence that all the inputs to STEM can adjust just fine to market signals if they're there. Training or importing a bunch of aeronautical engineers is not going to get us to Mars. Funding a Mars mission is going to get us to Mars.

re: "The current US immigration system already tilts the playing field towards the unskilled through family reunification and through the habit of tending to regularize or to tolerate the illegals"

That's an awfully loaded sentence. Family reunification doesn't tilt the playing field away from skilled migrants. Family reunification works the same way no matter what your skill level. We may end up getting more unskilled workers than skilled workers, but then again the world's population is more unskilled than skilled!

The only tilted part of the immigration system on the basis of skill is the work visas that explicitly deny the opportunity to get a guestworker visa to low skill workers and provide that opportunity to high skill workers.

If you want an untilted playing field, take all skill level qualifiers out of the immigration system.

Since you admitted that there are more unskilled than skilled workers and average immigrant quality is low and their families almost certainly lower than even those low immigratn averages, it makes sense to compensate for this. The immigration laws are not in any way market prices, so I'm comfortable saying that in equilibrium, the current system makes the unskilled more likely to be admitted than the skilled and that this would be improved by controlling the extremely loose definition of family reunification that's used.

I think aside from being cognizant that some people come here just to work (in other words, having a guestworker program), immigration policy should not be used to influence the composition of the workforce.

I think immigration policy ought to be about welcoming new Americans. Family reunification is great for that, although I suppose not necessary. Obviously we have a special interest in things like asylum too.

But I don't want Congress to game the workforce with immigration policy.

Why is that gaming? Isn't the workforce aspect a core part of immigration public policy by definition?

In average, it costs a skilled immigrant thousands of dollars in application fees alone and at least 3 years of waiting (can up to 6-10 years for Chinese and Indians) to upgrade his employer-sponsored H1-B visa into a green card. The sponsoring company has to demonstrate that his skills "cannot be replaced" by hiring American citizens. Anything happen to his company or his job in that waiting time frame, he will be out of luck and his "place in line" will be taken by others. Family reunion visas and asylum seekers on the other hand, enjoy priority processing of their green card applications and also a quicker path to citizenship.

Fact is the current system treats high skill, highly educated foreign workers no better than how the system treated Chinese railroad workers in the 19th Century. Again we see an agenda behind the system that purposefully punishes productive people who work hard to earn the right to "be here" while letting illegal immigrants and people who have no skill, no work and no education getting the green card and citizenship ahead of them.

I completely agree re: the lemons problem in grade inflation. I went to MIT, majoring in chemical engineering. I say this not because I want to brag, but because I encountered firsthand the effects of recent years' grade inflation in my own job search. While academia was cognizant of the fact that MIT grades still actually reflect understanding and effort well beyond the already solid effort required to attain the average, the private sector was not so forgiving. I had a roughly 3.15 average in undergrad (on a four point scale, not the five point equivalent MIT likes to use to salve its undergrad's egos) and that was enough to get me automatically rejected from most job interviews, despite being in an in-demand STEM field.

Later I went to graduate school at Cornell where grade inflation is not taken as seriously, and got straight As while also working part time and playing more video games than I had ever had time for in undergrad. Given that this was pre-recession, job offers came piling in when my inflated GPA came into play. All it took was an additional $40k in loans to buy the inflated grades...

Industrial policy by another name?

It all depends on when Kurzweil's Singularity arrives, doesn't it?

Kurzweil's been writing about that for 25 years now. His idea of a Singularity is looking a lot like Fusion reactors and flying cars.

Where's my flying car dammit!! :(

Yes, but the point is that the future is hard to predict, therefore it's a little tough to say that STEM should be encouraged. Nobody can possibly know if that's useful.

Well sure, but by that standard we shouldn't do any long term work or planning, because we don't know what the future will bring. That's a pretty useless standard.

So instead we make a best guess at what the future will bring. Kurzweil's vision doesn't strike me as a high probability event.

The point is that (IMO) the government should NOT be doing that sort of planning.

There is indeed a profound shortage of STEM workers, in much the same sense that there is a profound shortage of 2014 Corvettes on sale for $10.

The past twenty years has been dominated by the MBA and the JD. The same people who demand outrageous salaries on the premise that they are indispensible, seemingly have a difficuly time understanding supply-and-demand when it applies to other people.

If you are capable of getting a degree in a STEM field, then you are likely more intelligent and rational than the average person. And an intelligent, rational person is less likely to commit to years of graduate work given the low salaries and job security that seem to be the norm. Why work and sweat so hard, when your CEO is just going to send your job to India so he can get his quarterly bonus.

When STEM grad students can expect $100k job offers out of the gate, and MBA's have to live with their parents to make ends meet, I bet our "shortage" of STEM workers vanishes rather quickly.

(Have both a MBA and most of a Ph.D. in physics. Gave up the Ph.D. after I met brilliant people in my field who were in their 10th year as a postdoc and needing food stamps to make ends meet.)

I have a STEM degree and so do many of my peers - even recent grads I am acquainted with from my very STEM-heavy alma mater are having no trouble getting job offers. Usually they have some internship experience in addition to their degree. They may not be getting $100k but they get into the $60k range pretty quickly. They have experienced layoffs but few lengthy unemployment periods. They also regularly receive recruiter calls to lure them away.

I am also a hiring manager in a STEM field - software - and have a devil of a time finding candidates, particularly at the higher levels. And those we do find, we are often in a bidding war with another employer.

For a reasonably high tier college (let's say top 50 or 100), $60K is a pretty marginal wage benefit. That's lower than what the majority of my peers as an econ major started at.

I'm the CTO at startup in Silicon Valley and have been a hiring software engineers for years. It's always been a struggle to find *qualified* candidates.

But there has never been a shortage of applicants with degrees in Computer Science. The majority of CS graduates are significantly less skilled than say, the top 25%. Thus, all tech companies that know how to identify the best candidates are competing for that small slice.

The rest of the engineers out there work for some of the larger orgs where they can't properly vet engineers and are stuck hiring whoever they can. Unfortunately, there aren't enough jobs in the market for such candidates to have a real measure of security in their careers.

So you can argue that we have both a shortage and a surplus of STEM graduates. But really, there are just a lot of poor STEM programs that either offer insufficient preparation or are not challenging enough to weed out those people that aren't capable enough for an actual career in a STEM field.


The PR machine says: Can somebody else, like the government or anybody else, please create incentives and remove disincentives for people to go into STEM, so that we don't have to pay quite so much?

I would also add that a top-level Masters grad from a top tier STEM school can probably quite easily get $100k offers.

Any Masters of Computer Science grads from MIT or Stanford here to comment?

It depends significantly on location. If they are willing to go to NYC or CV they can easily break six-figures fresh out of school, but they will have to pay the higher living expenses.

A lot of people with masters from MIT go into business or finance, though, because the pay is better. You could make them into engineers by paying them more.

CV was supposed to be SV, Silicon Valley.

Hello! My degree isn't a Master's but it is from Stanford and is in Computer Science. I've been in the software industry for 26 years and have been wondering for 25 of them when the generous salary I get paid will fall in line with what other bright, well educated people (such as my siblings and cousins) make. It is happening for lawyers in the USA because supply has finally exceeded demand, which I would have thought impossible, but nothing of the sort has happened to me.

My conjecture is that there are bright people who aren't very motivated by money. Anyone who gets a Ph.D. in astronomy, for one. They probably see the salaries in my field and think, "Not worth it," just as I saw the higher salaries on Wall Street (higher than mine, I mean) and thought, "Not worth it."

What's an economist to do when greed fails?

Ever tried working in the oil industry? They pay well to engineers. If your gut reaction is disgust, you have a problem of taste but the STEMs vacants paying 60K on your first year are there.

Half of Ms and Ph D students in the university I've met end working for mining, oil, environmental or energy industry. The other half get into consultancy, get tenure somewhere and a sad minority starts a life complaining about the uselessness of STEM grad school,

Ps. It's up to you to monetize your grad school knowledge. No one is going to do that for you.

Petroleum Engineering wages and employment are very cyclical. In boom years they are the highest paid engineers. In bust years it is very hard to get a job.

That's why people move around the world running away from busted fields


Grouping STEM together as one category is highly misleading. All of the people I know in the T and E are doing great, but those in the M and S categories aren't. I'm in the biosciences at Stanford and the majority of people I know end up in endless post docs and then usually get jobs that were no where near what they trained for and that pay far less than would justify obtaining the PhD in the first place. Binkley is spot on here that employers simply don't want to pay for the higher level skills they require and the constant overproduction of STEM workers (to say nothing of immigration policies) allows them to constantly depress wages. My personal observation from working at three different universities during undergraduate, technician, and graduate work is that only 25-30% of biosciences PhDs ever get science jobs and a comparably small fraction make the 100K+ on a reasonable time frame necessary to justify the investment. If you make no money until you're 43 and then start at 90K a year, your life time earnings aren't going to be much better than if you hadn't bothered going to graduate school in the first place. It's all well and good for people to say "Hey, why don't all of you work for 12 years or so, 6 or 7 days a week, so that you can get a STEM degree. Sure, it won't pay off for you, but think of all those positive externalities!", but that doesn't mean that we should keep pretending that there's a shortage or keep trying to convince young people that it's a great investment.

Are we talking about PhDs? Are PhDs really necessary or just regular B.S.?

I'm talking about PhDs for the M and S parts of STEM. A four year degree in biology or chemistry doesn't pay all that well compared to the effort necessary to obtain it. Starting technician salaries are around 30-40K, and they top out around 50-60 later on. You could easily make that much studying poetry and teaching high school, all while getting 3 months off a year and dodging calculus.

It's my impression that a BS is sufficient for many IT and engineering jobs OTOH.

Don't we need to distinguish between fields where the only job prospects are academia, and everything else? Plenty of biology PhDs, I assume, get paid well in biotech, medical devices, bioinformatics, finance, and whatnot. But in my opinion, going into academia is often a pretty poor career choice these days. And that's before the entire field collapses as a result of online education and various other trends.

Plenty of biology PhD's also end up working as glorified lab techs in never-ending "contract" positions. My last job put out an ad for a a $32k/yr lab tech position and got people with Ph.D's applying.

Pretty much this. The company (well, the business unit of the company) I work for does some real bleeding-edge work, and we've been desperate to find qualified people for a while now. Mind you, we've interviewed a number of good candidates, we just haven't been able to hire them because HR won't let us offer a salary they'd actually accept.

Now we're getting a flood of new managers, who all seem to be the low-rent MBA program grads that find no problem with the above. They instead insist we hire cheap contractors. Never mind that the contractors can't actually do the jobs they're being hired for, but existing qualified people have to take time to hand-hold them through their work (or flat out do it for them so we don't have to wait for them to mess it up before fixing it). And despite not being able to offer suitable salaries on the engineering side, the managers sure don't seem to have a problem with buying new houses and (it at least one case) Aston Martins.

I should add that there are obviously firms out there that don't have this problem. None of the people we wanted to hire had a problem finding positions elsewhere at higher pay. But I happen to know that the BigCorp that we're working for is one of the ones publicly complaining about the supposed shortage.

Or the perceived shortage is PR whipped up by the employers who would love to lower salaries they pay from 6 to 5 figure ranges, without losing the supply of skilled labor. Not that's necessarily such a bad thing.

This post has more serious consequences than Prof. Cowen may think. Even if all this is true, it suggests that policies to favour STEM workers (e.g. preferences in immigration or in education funding) will not be good for growth. Instead, they will primarily benefit incumbents who seek to hire STEM workers, unless society also makes the extra investment in particle colliders. But if neither increasing workers and investment are happening among private capitalists today (at the margin, to the socially-optimal extent), this becomes a theory of state-led industrial development. Unlike South Koreans, we do not know what the outcome looks like. It's easier to produce auto workers than to produce whatever skill it is that we may need at the margin in fifty years' time. So perhaps we end up with the Keynes-Cowen theory of capital: private capitalists don't allocate investment well and we need centralised decision-making.

I would be interested in a study of high skilled immigrants returning to their home country after being trained or employed in the US.

Just looking at inputs (immigration) without looking at export of talent (emigration back to the country) is a weakness.

How many electronics, software, and investment banking firms send their trained, new US immigrants, back to their home country to establish a branch, set up a new factory, etc. for items or services that would have otherwise been performed in the US.

I haven't had the time to read the paper yet, but there's a lot of NSF money in alleviating the gender imbalance in STEM majors. So I was just curious if this particular analysis also suggests that there isn't a shortage of women in the STEM workforce as well.

Great, I was just getting my head around that this was really an immigration issue. I'd like more STEM for the exact same reason I want my kid's elementary school to focus on read, writing, and arithmetic.

I would think that the party of science would want the baseline education to be scientific. If they did, I'd have to re-think my tendency to hold that view.

It depends on what STEM students will be doing. Are they workers or innovators? The latter involves the risk of funding research that may lead to a dead end. The money for that has to come from somewhere. For South Korea, there's a lot of public funding as well as the profits that come from the market power of the Chaebols.

I think one could argue that the fall of great US innovation laboratories like Bell Labs was related to AT&T's loss of monopoly profits. PARC did amazing things, but it got money from NASA, DARPA, etc.

There aren't too many companies willing to throw money at projects that may never be profitable. I think any new round of innovation will require public funds, either through gov't contracts, tax incentives, or grants via NSF, etc. In this era of sequestration and politicians taking pot-shots at research projects with funny names, I don't see that happening.

I bet computer science students could get jobs coding the next Angry Birds, or chemists could get jobs developing new artificial flavors for Doritos, but innovation requires some "foolish" spending on far-fetched ideas. Google does a lot of that. I think one could argue Jeff Bezos at Amazon does too. In a sense, for innovation, you want foolish spending, and that's one thing the government can actually be good at.

Is manufacturing high precision bearings in increased volumes in the US a high risk proposition when the market demand is well established and huge? That requires hundreds of STEM workers in manufacturing, skilled "real" machinists, manufacturing engineers and technicians, quality engineers and technicians,... The time required to produce such workers is on the order of five years. This has been an ongoing issue in New England for a decade. US manufacturers sold out to Asian and European manufacturers because they could not solve their STEM worker problems, and the foreign manufacturers seem more interested in working with government to get the STEM workforce built they need.

By the way, this is something the South has done - work with Asian and European manufacturers to train STEM workers for free to meet their needs to get the plants built in their States. I read recently of a "win" which creates two thousand jobs paying less than $100K including benefits at a cost of taxpayers in excess of a million dollars a job, with the local, State, and Federal government paying to train the workers to the Asian employer's specifications.

Here in New Hampshire the Republicans are different and won't tax and spend lavishly to create jobs like Republicans in the South.

I wouldn't call that 'foolish' spending, more like spending that doesn't have near term profits as an immediate and most important goal. And yes, that's one thing the government can be good at. Because it's not a private profit-maximizing entity, it can spend funds on basic research just to see what gets discovered. History has shown this to be a positive function of government spending, which you mention (Bell Labs, PARC, etc got a lot of govt cash).

I believe Steve Jobs and Tim Cook discussed the core problem with the lack of STEM workers in the US.

In no place in the US can you find a large number of the following STEM workers:
- design engineers
- manufacturing engineers
- manufacturing technicians
- manufacturing process engineers and technicians
- quality engineers and technicians
- manufacturing supply chain specialists
- manufacturing supervisors able to quickly organize and train workers

For a corporation like Apple, their product strategy requires a mix of these STEM workers by the ten thousands within 60 days.

That requires a pool of such workers within a zone that numbers in the hundreds of thousands so a large number can be recruited quickly.

But to have such a large pool of such workers requires a large and thriving manufacturing industry at critical mass.

What much less diverse operations requires make it next to impossible to manufacture in the US. For example, New England is a center of bearing making that goes back the region's contribution to the early wars of the US. This requires experienced machinists with knowledge well beyond lathe and milling machine operator/programmer. To setup new production in NH requires 100-200 real machinists, but that is 40-50% of all the machinists in the region who are already fully employed, and who have an average age over 40.

The US no longer has any US bearing manufacturers as they have been bought by Asian and European based manufacturers, so expanding their US plants comes down to simple tests: can we find 200 machinists in NH, or should we pay the very high wages machinists demand in Japan? The answer keeps ending up "we pay the very high Japanese wages" or probably "we pay the lower Korean, Taiwan, China wages". New Hampshire does have a machinist tech school and a machinist apprentice program with business support, but these programs can barely offset the death and retirement of the region's machinists.

Economists largely ignore time in all their theory. It is like, if I raise the offered wage by 10%, the 5 year training for an machinist will become 4, so a 50% wage increase will produce a hundred machinists with 5 years experience within a month. And in New England, a high percentage of machinist run their own shops in large part due to the large corporations cutting them lose in the 80s so they will demand long term corporate commitment that few economists these days argue businesses should offer. Anyone arguing the key to increased STEM worker supply is life time employment with robust pensions?

Timken is still a US company, with operations in New England.

Apple's contract manufacturers employ tens of thousands of workers because they are available cheap. It isn't 'required' by the nature of the product. Smart phones could be manufactured in a high-wage country, but with far more automation (both hard and flexible) and far fewer workers. The New York Times didn't understand this either.

What a surprise: nominally libertarian person of note retreats from "the market will handle it" on one of the rare occasions when labor actually develops some power against capital in said market.

If you want more STEM grads, you need to pay more money for them. Period. Importing them from other countries to work as indentured servants is gaming the market as much or more than if Obama forced every industry to unionize.

No, it isn't. That's like saying globalization and free trade is the same thing as tariffs and subsidies.

Not even close. H1B's aren't free trade; they're artifical depressants for the price of local labor. A closer analogy would be if Apple didn't like the price they were paying for some US company's product input and they got the government to subsidize cheaper competition from abroad for them to buy.

I'd add one caveat here, which is that there's a non-trivial number of science workers (grad students, postdocs, etc) whose salaries are set by the government. In theory, I suppose one could try to negotiate, but in practice, almost no one does for trainee-like positions. The situation is probably different for principal investigators at national labs, etc.

Another viewpoint on this issue as it relates to hiring software engineers in tech companies is that, on average, careers in the field are short, so fresh meat is always in high demand. Fresh dumb meat.

When I was young, I made a calculated choice to go to law school rather than grad school for computer science. If the US starts drawing in more STEM graduates from the rest of the world the rational response for American students will be to reduce enrollment in American STEM programs.

Now personally, I think that the United States has a critical shortage of competent economists and that the US should be encouraging immigration of foreign economists who understand the distinction between correlation and causation and have advanced skills so they can correctly use the copy/paste function in Excel. (The United States also needs smarter Congressmen but they have successfully protected themselves from foreign competition.)

We also need to import foreign executives, who are much less expensive as well.

One of the problems with using wages as a proxy to attempt to measure worker shortages in the IT industry is the relative inflexibility of software companies to adjust wages to respond to their own internal labor shortages, and to give pay raises internally to their better workers.

I work in the software industry, and I've seen companies pass on qualified workers--and leave job requisitions open for many months--rather than pay the extra 5% or 10% the worker asks for.

Part of the problem is inherent to the software industry: because a qualified worker can be several times more productive than an unqualified worker--and in some cases an unqualified worker can actually create negative progress for a software company by writing code so bad it has to be ripped out and rewritten later--that software companies are hesitant to hire and pay exceptionally well qualified developers a higher wage for fear that they're hiring a con-artist instead.

So when there is a genuine shortage, software companies are loathe to raise salaries, out of (what seems to me) a habit of capping salaries to avoid paying excessively for what turns out to be an unqualified individual.

As a proxy to software salaries, I would suggest instead studying the number of workers who leave large companies to go out on their own. One reason why I struck out on my own is because the demand for my services (I develop iOS and Android apps) is so high, I'm able to easily find contracting work--and thus I'm able to capture more of the gains of my own productivity through greater intangibles (more work flexibility; since striking out on my own I've had the time to go get a pilots license), and through bidding work 'fixed-price' and completing the work faster than most.

I know there is a coherent argument in here somewhere. But I think it's this:

(A) Software companies generally use a fixed formula to determine salaries, and those formulas are relatively inflexible compared to market conditions. Software companies sometimes will even punt on large scale (and potentially very profitable) projects because they are unable to staff them--even if the rewards from those software projects are potentially great enough to justify paying their staff a larger salary.

Because of this,

(B) As the labor shortage increases the gap between what a software developer could earn as a freelancer verses what they could earn salaried grows, and this encourages software developers to realize those gains by striking out on their own.

As a consequence,

(C) I always take comments by places like Google or Microsoft about an inability to hire qualified developers with a grain of salt. The problem is not an inability to hire; the problem is that large companies don't want to pay at the price point the market demands, in part out of fear of hiring the wrong people.

Interesting points, William. Do they amount to saying that software development doesn't scale?

I used to be part-owner of a small software compant. We never had more than four programmers, but sometimes we still got stuck with people who looked good on paper and interviewed well but wrote very bad code.

We don't all agree on goals, so our recommendations are in conflict. If the goal is reducing US unemployment, then H-1B is clearly not the way to go.

If the goal is increasing the US GDP, then true immigration of highly-skilled workers is desirable, and H-1B is also not the way to go.

If the goal is increasing the profits of IBM, Microsoft, Google and Apple, then the H-1B program is exactly the right solution.

The US system emphasizes the high-intellect roles of the most advanced thinkers and workers. Germany has an excellent industrial and employment situation by also emphasizing (and culturally valuing) the role of machinists, manufacturing engineers and other specialists who are not seen as Einsteins, but as manufacturing specialists.

If the existing supply of STEM students is greater than the available positions, a large minority of those positions face substitutes from non STEM students (about 40% non college and non STEM major from college ?) and businesses are more interested in hiring smart, creative non-STEM graduates over STEM graduates (and training on the job is then implied) what is the marginal value of another STEM educated student?

Even with some assumption about positive externalities it's hard to see why those externalities are not present for the not STEM candidate that is hired so why more STEM?

The argument doesn't seem to hold.

"we do not observe significant wage rises for STEM workers"

Could it be that the 'urgent' need for STEM workers is based on a theory of a technology based economy that has yet to be proven or materialize? Not saying the theory will not prove to be sound some day, but as of yet it is still a theory that goes something like 'wouldn't it be cool if we had thousands of Apple's and everyone were a techie making big bucks working for them, instead of doing drudge work in factories or service jobs?'

I don't know about STEM workers, but we definitely have too many STEM students because they disproportionally blow themselves up: - When was the last time an economist or a historian turned into a terrorist?

Is investing in auto capacity ever a good idea, in Korea or anywhere else? World is not suffering from any shortage of auto capacity. How many years have GM and Ford earned cost of capital? Korea is one small market served by Kia, Hyundai, Ssangyong, Daewoo, etc etc. Korea and the Chaebols are horrendous at allocating capital.

Is it worth pointing out that there is a 55-year-old (!) paper by Arrow, Alchian, and Capron that addresses this question?

Wow! Isn't it amazing to see just how far the field of economics has advanced in the past 55 years?

I think we need to stop talking about STEM workers as a monolithic group. There's a big difference between someone with a degree in Petrochemical Engineering and one with a degree in Environmental Sciences or a general degree in biology. The shortages tend to be very specific, but we insist on talking about STEM as if it was all the same.

tougher schools? Really?

there are 3.9 million jobs unfilled in the USA as per .. why can't these openings be filled by the unemployed? if the currently unemployed cannot fill these gaps, who will?

It seems to me that the failure of wages to rise doesn't signal a shortage of STEM talent (locally). That wages haven't risen is a sign that there is adequate supply of that talent elsewhere. Increasing the supply of STEM talent locally allows more of those high-quality jobs to stay here instead of migrating oversees. It changes the skill-mix of our population, and, as Andy Grove has argued elsewhere, enables more innovation here going forward.

Comments for this post are closed