The United States is Starved for Talent

The US offers a limited number of H1-B visas annually, these are temporary 3-6 year visas that allow firms to hire high-skill workers. In many years, the demand exceeds the supply which is capped at 85,000 and in these years USCIS randomly selects which visas to approve. The random selection is key to a new NBER paper by Dimmock, Huang and Weisbenner. What’s the effect on a firm of getting lucky and wining the lottery?

We find that a firm’s win rate in the H-1B visa lottery is strongly related to the firm’s outcomes over the following three years. Relative to ex ante similar firms that also applied for H-1B visas, firms with higher win rates in the lottery are more likely to receive additional external funding and have an IPO or be acquired. Firms with higher win rates also become more likely to secure funding from high-reputation VCs, and receive more patents and more patent citations. Overall, the results show that access to skilled foreign workers has a strong positive effect on firm-level measures of success.

Overall, getting (approximately) one extra high-skilled worker causes a 23% increase in the probability of a successful IPO within five years (a 1.5 percentage point increase in the baseline probability of 6.6%). That’s a huge effect. Remember, these startups have access to a labor pool of 160 million workers. For most firms, the next best worker can’t be appreciably different than the first-best worker. But for the 2000 or so tech-startups the authors examine, the difference between the world’s best and the US best is huge. Put differently on some margins the US is starved for talent.

Of course, if we play our cards right the world’s best can be the US best.


"Of course, if we play our cards right the world’s best can be the US best."
It worked so well for Rome.
Listen, we want less immigration, not more. We want less trade, not more. We want more jobs for Mericans, not less.

Less trade, more jobs? Good luck with that.

I obviously meant imports, not exports.

More mercantilism, more jobs? Good luck with that.

Many countries got rich(er) by managed trade: Red China, Japan, America in the 1860s, most of Europe.

Neat trick! I, too, can deploy rainbow ruses in support of my ideas.

Are you familiar with Plato’s dialogues?

The best markets to sell information products to are: (1) Business Owners Seeking Solutions (2) Better Appearance Seekers (3) Business
Opportunity Seekers (3) Diet & Fitness Seekers (4) Dating Advice Seekers and 5) Avid Leisure Hobbyists.
The Washington lottery site is well designed and very user friendly.
To get you started, repeat over and over again to yourself,.
It is a drawing game, and it pays out the largest
jackpot of any Vermont lottery game.

Starting with the early Republic, Rome built its strength with the first foederati, called socii. Yes, 600 years later other foederati caused trouble, but Rome could never have become what it did based on the population of the one city.

"Despite these benefits, many socii rebelled against the alliance whenever the opportunity arose. The best opportunities were provided by the invasion of Italy by the Greek king Pyrrhus from 281 to 275 BC, and by the invasion of Italy by Carthaginian general Hannibal from 218 to 203 BC. During these invasions, many socii joined the invaders, mostly Oscan-speakers of southern Italy"

Let us be blunt: we have enough pobkems and do not need to import backstabbers.

H-1b visas are not about talent. They are about money or profit. H-1b workers usually replace workers in fields where 80 hour weeks are expected and demanded. American workers not only do not like that kind of slave labor but they also insist that if they must work those hours that they get paid for it. H-1b workers will work those hours for half what American workers are paid. THAT is the sole reason for H-1b.

Any evidence to back this up? My experience is that all of the people I know who applied for H-1B visas were graduating students who immigrated to the US for a highly technical higher education (MS/PhD) and now wanted to stay and work for a US company. From where I sit, I see plenty of job options for these students and their American counterparts. And, I don't know that many that work 80 hours a week. If the smartest, hardest working people in the world want to come to the US, I say let them come!

Only anecdotal evidence, but I have quite a few friends working in the financial services sector on H1-B visas. Because of the sponsorship and job requirements, they cannot freely move from one employer to another. This gives incredible power to the employer in dictating the financial conditions of employment, as they know the odds of the employee leaving for a better offer are much more limited. Once these employees get their green cards, they all celebrate and leave for better offers from rival firms.

I don't believe this is the way it works. Actually discriminating within the firm would require low-level managers know a lot about it and never defect and blow the whistle or participate in a lawsuit.

If I was an evil CEO and I wanted to do something like this, I'd hire a lot of H1-Bs and some regular Americans and I'd treat _everybody_ the way you guys describe, with slow promotion, slow pay increase etc. And I'd let the Americans sort themselves out by leaving for better opportunities.

Saul is correct. I was software engineering for a long time. I have hired hundreds of people. I can assure you that h1b holders are underpaid virtual slaves hanging on for a green card. A green card allows them to move from one employer to another, the H1b does not. These young single men know they have to work like slaves, working overtime on the cheap, or they will be dumped. You will see some talented green card holders, because they are the survivors of the H1B game.

All of this hurts American citizens. It discourages people from entering STEM careers or studying STEM in college. It also allows companies to discriminate against older workers and women. As I said before, the vast majority of H1b holders are young males. The program is inherently discriminatory. It should be junked.

It's all about the Benjamins.

H-1Bs can move from one company to another. There are constraints on the movement. Mostly they can move from one big employer to another big employer.

As far as I can tell, outside of a handful of visa mills, H-1Bs are hired into ordinary college-grad jobs. Because they expand the applicant pool, and I acknowledge because it's somewhat harder for them to move, they drive down the prices firms need to pay to fill those roles and they drive firms to invest less in productivity improvements.

Overall, it's a bad program that needs serious changes. But, it's better than what goes on with immigration in the rest of the economy. At least these folks are employable and paying taxes.

It's really much worse than just the wage suppression via added supply. The H-1B holders are typically only taking the jobs at the offered low wages in order to eventually get citizenship. And then the goal is to bring over their extended family. They talk about this openly on forums you can peruse. The private employers are essentially auctioning off US citizenship and pocketing the proceeds.

If we should have "guest worker" visas at all, there should be absolutely no path to citizenship or the option of bringing over your siblings and cousins. You can come here and work, you and your family will never vote.

It’s both.

Increase HB1 visas to 300-500k a year and make them industry specific not employer.

And let them apply for a green card in 12 months.

If we want to let someone live and work here, we should issue them a regular green card and let them work where they will. Chaining them to a given employer makes no sense in terms of economics or fairness, but makes political sense.

If US firms are really starved for talent, and H-1b visa holders really add so much value to firms, firms can afford to pay large sums of money to the government for a visa and the H-1b visa holder should be able to leave for a market salary after a short period of time. But you need firms to pay large sums to make sure that the talent is really what they are after, as opposed to undercutting existing wages (of both US citizens and resident aliens).

If only you could actually pluck the best of the best and not settle whole nations on your territory. If only modern society provided some way to know IQ, or maybe filter for high education, or maybe even know abilities from specific achievements in the certain fields, like scientific papers, or competitions in programming. If only we lived in a society, where we could not equate unqualified proletariat from highly qualified specialists. If only...

By the way, I am not from USA, so I am 100% for you to not accept any good specialists, our competitiveness will grow.

Thank you, but I don't want invaders. America intends to remain the master of its own house. That is why we fought in Concord, St. Mihiel, Guadalcanal and Suoi Bong Trang.

. If only modern society provided some way to know IQ, or maybe filter for high education

Why aren't businesses responsible for the education of their workforce instead of being subsidized through public spending on secondary and higher education? There's no reason a tech-oriented business can't search the country for potential employees, educate them or pay for their education and then put them to work. It isn't necessary for there to be an enormous and expensive apparatus to supply these businesses with effective help. After all, in the normal course of events the public doesn't give them real estate, buildings, raw materials and production machinery. Of course as we devolve into more and more of a socialist society and culture it does seem as though it's up to the general population to provide for the needs of commerce. Maybe Walmart should just purchase Harvard University and use its graduates for its workforce.

I think all of these comments betray a lack of understanding about the way the current system works. (1) We don't have to filter or IQ, they come to the US university system (the best in the world) because we provide the best opportunities here and abroad. And then, they come and pay non-resident tuition or work for very cheap doing basic research as a graduate student.

If that's indeed the case, and it probably is, many of the foreign students are in the US because they come from families with money, not that they're necessarily on the highest end of the intelligence curve or ambition meter. They can afford to pay full tuition and work for cheap since they already have money.

For undergraduate students, yes. For graduate students, my experience is that this is not so. Most technical graduate degrees in the US come with stipends of $20K or greater. Most single people can live on this well enough in their 20s.

Because you get to carry around your human capital with you in your head, and sell it to the highest bidder. Why should businesses pay to educate workers so they can go work for their competitors?

If you pay to educate them they might work elsewhere. True. But what if you don't educate them and they stay put?

You sure it isn't someone else?

“Invaders”, misspelling problems with a k, the intentionally irrelevant list of battles...

We fought in Vietnam to remain the master of our house and keep immigrants out?

Classic Thiago

The H1b is a fraud. Most H1b holders are very average but cheap young unmarried men. Most of the h1bs are snapped up by the giant Indian outsourcing companies like Wipro et al.
Even when US companies use the H1b to bring in "talent" the main purpose is to drive the price of labor down. Given that these visas are usually used by high tech companies where most of the money goes to the founders and early hires via stock options, it is wholly unnecessary to use the H1b to find mid-level skills like the gaggle of average software engineers we bring from India. The greedy bastards that run those companies don't want to share the money with the fellow citizens that actually create and support the environment in which their business thrives.

We find that a firm’s win rate in the H-1B visa lottery is strongly related to the firm’s outcomes over the following three years. Relative to ex ante similar firms that also applied for H-1B visas, firms with higher win rates in the lottery are more likely to receive additional external funding and have an IPO or be acquired.

Since the vast majority of H1-B visas go to large established and already public companies, I'm wondering if this is a bullshit metric. If only 3% of the H1-B visas are going to startups then a 23% increase in IPOs isn't all that impressive in the grand scheme of things.

Although, if you're suggesting that public companies should be excluded from the H1-B program, then I'm certainly on board with that.

> Relative to other firms that also applied for H-1B visas

iow no sample bias if assignment is random.

John is absolutely right here, with the exception that limiting the ban on H1-B to just public companies wouldn't be sufficient. Two of the top five H1-B sponsors are privately held accounting/consulting firms (EY and Deloitte).

"Top technology companies like Amazon, Google or Facebook are the most prominent seekers of H1B workers in 2018."

Still, Tyler's overall point that the U.S. is starved for a very specific type of (computer) talent remains true.

Bullsh*t! The USA, a nation (for now) of 330,000,000 people with free access to the best public education system of it's size on the planet, does not have a skills deficit. These a$$holes at the FAANGs and their future acquisitions don't want to share their unprecedented prodigious wealth.

Don't be played for fools.

If the money these bastards are hoarding were spread around a bit we would have a larger middle class. A lot of money would be in motion rather than sitting in a stock portfolio.

We have all been suckered into walking into our own death Chambers.

Someone here with a calm, rational, not at all hysterical take on the situation.

Death Chambers of Commerce. Good one.

"These a$$holes at the FAANGs and their future acquisitions don't want to share their unprecedented prodigious wealth."

They are so greedy that working at any of those companies easily puts you into the top 1% of the US.

If they weren't making even more than what they pay they wouldn't pay those salaries, so they're not doing it out of the goodness of their hearts. I'm not saying that they are or aren't greedy, but you can't use the salaries they pay as an argument against them bring greedy, after all they can be and still pay high salaries, two things can be true at the same time

"the vast majority of H1-B visas go to large established and already public companies" perhaps because large companies already have staff on-board with plenty of experience doing this, and small companies don't?

One, see the public and higher education apocalypses. Two, America is starved for talent that will work for cheap.

Throw more money at phony education and open the immigration floodgates! Brilliant.


"...will work for cheap."

You're cutting up some free-range truth there butcher.

"America is starved for talent that will work for cheap"

How do you reconcile that with flat wage growth, even among the middle economic quintiles?

He's right, though.

An American student might be talented, but isn't going to invest in turning that talent marketable if the return on that investment is abysmal. Flat wage growth can occur if companies just wait for the right H1-B worker they can pay peanuts for. The American talent worker can compete for those same peanuts, or find something else to do with their lives.

The return on college education is still in the neighborhood of 14%.

First, I cannot reconcile why we have 330 million people in country and 'they' need to import talent.

If I was so inclined, I'd identify over what period and what market segments suffered flat wage growth. I'd try to look at supply and demand and multiple-factor levels and trends.

I apologize. I'm past reconciling big stuff/data. I retired from my microscopic, micro corner of that world. Doing nothing [and baby-sitting grandchildren] takes up so much time. It's a huge challenge to fit in nap time.

You can't look at broad aggregates and make statements about a subset of the labor market. Wages for tech talent have grown pretty substantially. Many companies are also establishing tech centers in Asia and Eastern Europe because there simply isn't enough talent in the US.

Or brcause they can hire people for lower wages.

The lower wages aren't as beneficial as you'd like to think. US talent is far more productive and far more creative - it takes *at least* twice as many people overseas to get anything done and the quality simply doesn't hold up.

H-1B requires that you pay prevailing wage. This isn't a cost play.

In fact, sponsoring H-1B is expensive - we try to avoid it when we can due to the costs

"H-1B requires that you pay prevailing wage. This isn't a cost play."

The first part is true, but the second part isn't. You double the size of your applicant pool so you don't have to raise wages or invest in other productivity enhancement to get the results you need.

that's a nice assertion you have there - but as someone that has been involved in hiring H-1B for a long time I can tell you:
1. That it is more expensive to hire H-1B than someone that is already here due to all of the associated fees and general expectation that we'd sponsor green card as well.
2. We'd much rather hire and fill someone *now* as opposed to much later
3. The size of *my* application pool is irrelevant to prevailing wages no firm really has enough market power to set prices like that.
4. Before you say "but Microsoft" (or Google or Amazon or...) they are already paying better than almost anyone else.
5. Even ignoring ALL of that - there aren't enough H-1B visas to make an impact in any labor market.

I don't know what to tell you. I've been a hiring manager for H-1Bs for 20 years. I don't work in HR.

1. The legal costs are small in the scheme of things. You're mostly talking about college grads here. The risk that visas won't get renewed has become a real thing, though it's not huge.
2. We mostly hire out of school, so there's no difference in timing - it is different with the visa-mill IT consulting employers discussed by others below.
3. The size of the applicant pool has a direct effect on the wages I need to pay to get the yield I need to have to staff my projects. Half my applicant pool, and I need to offer more to get twice the yield. That's not complicated.
4. I don't know what the "but Microsoft" objection is supposed to mean. Most H1-Bs are going into medium to large businesses.
5. Sure. The H1-B drives down wages for middle-of-the-distribution college grads in mildly in-demand fields like STEM or econ. It has no effect on blue collar wages, for example.

My experience is that for students graduating in the physical sciences, getting an H-1B visas makes it significantly more difficult for a non-US citizen to get hired than a citizen. I talk to more people like Chris than like Lord Action.

I don't think your perspective is inconsistent with mine. An H-1B limits you to employers large enough for it to be worth the hassle. So it's an impediment to employment. Amortized over 20 h-1Bs, though, it's not a big problem _for the employer_.

Also, the physical sciences are not a great point of comparison. You hear people talk of STEM and even STEAM, but really it's just TEM, or maybe TEM plus econ (which serves as compsci-lite with a handful of data science).

In any event, I'm not talking about the perspective of the potential employee with an H-1B. That's not an important factor in the design of the H-1B system. The H-1B system is supposed to operate for the benefit of Americans.

Just a point of clairfication: when I said mean physical sciences, I meant not CS/information technology. In other words, all engineering, chemistry, physics, materials, etc. (I am a Chemical Engineering PhD). I think there is a reasonably big difference between companies that do things with molecules and companies that do things with (only) code... Sorry, in retrospect not a useful distinction by the shorthand "physical sciences."

Thanks, that's helpful. And I agree that needing an H-1B is an impediment for the applicant and it limits the set of employers he or she can realistically engage.

But that's all a separate issue from the employer-side effects and the net effect on the economy and on the people competing with H-1Bs. If anything "it's harder to get a job as an H-1B" means they'll be willing to settle for less from employers, and that in turn drives down the wages of Americans competing with them.

With respect to talking to people with Chris, I think that's the point of the system. You don't want to have to put a person like Chris in a position of having to scheme and consciously discriminate against Americans, in part because they'll spill the beans. They need to have a simple story that makes their actions seem reasonable. That story is "we treat everybody the same way."

That may be true when the offer is accepted, but once that occurs, there is no market incentive for firms to continue that going forward. All of my colleagues on H-1B visas complain that they are not promoted or given raises commensurate with non-visa employees. This is because it's difficult for those on visas to leave their current employer for a rival firm.

Exactly. The employer dangles the green card under the nose of the H1b holders and thus has a virtual slave until they decide to dump them. I had plenty of h1b software engineers on my staff (not my choosing) and the following applied to most all of them:

1. They were young, unmarried makes.
2. They hung around and endured almost any insult hoping for a green card.
3. Their skills were average. None was a "super star".
4. Most never got a green card afaik.
5. Many were layed-off and got on a plane back to their home company. A few were actually needed, like the build engineer my retarded and now defunct employer riffed (by a retarded politically motivated Mangler). As a result, we almost couldn't build the product anymore. Totally stupid.

6. They were payed less in general, significantly so. I was "given" one guy from China that was paid $75,000/yr when most of my team was payed ~ 120,000/yr. Not a big salary in Silicon Valley, sadly. This case was the only one I ever even heard of being investigated.

One issue is there is no significant enforcement of the law, and the law is so vague employers don't even need to cheat. They just post a job somewhere in the media, gather a bunch of candidates, and then select what they believe ( often in error) is the best deal. The H1b candidate, by virtue of their lower cost, has an advantage. Nobody challenges the employer to prove the H1b was more qualified than the other candidates.

The whole thing is bullshit.

The only way to put a brake on this charade is to reduce or eliminate the number of visas.

"The only way to put a brake on this charade is to reduce or eliminate the number of visas."

Auction the visas. If the H-1B cost the employer $100k/year or $1m/lifetime, H-1Bs would be used in situations where there's a pressing need to hire that specific person and an American couldn't reasonably substitute. You'd remove the cost advantage from the equation, and firms that really needed international talent could easily get it.

If I recall correctly, Microsoft responded to the visa crackdown with a Vancouver research center. Does anyone know how this kind of off-shoring is going?

I'd be on board with more visas as well, but not this year, eh?

Offshoring? Microsoft Vancouver went from 400 employees to 500 over the span of a few years. Did you bother to read the link?

For reference, Washington state has about 45,000 Microsoft employees.

The link was for convenience. My question is for knowledgeable readers.

Option 1: Microsoft claims to be an American company and has access to Federal contracts and bailouts, while it hires foreigners over Americans.

Option 2: Microsoft moves overseas and ends the charade.

American workers and voters would probably opt for 2.

This is something a lot of large tech companies are doing nowadays: Typically they've hired the H1-B candidate already as an intern or as part of practical training, and if they are any good, they absolutely want to keep them. Opening an office in Montreal or Vancouver, if not both, and hire Canadians, along with those unlucky with the H1-B is far better than throwing the 1 to 2 years of tenure they gained during the F1.

It's also important to note that there are two different blocks of H1-B candidates: One wants a visa to come here (typically from India), and plan to work with a consulting firm (as going this route is far safer given intricacies of the later green card program), and then there's those that were international students with a US or Canada degree. The first group just keeps retrying, and they are not worse off if they fail. The second group, the one most high end tech companies and startups pick from, were already here, with a great US education, a good work ethic (as the weak performers are not going to get an H1-B sponsorship), and if they don't hit the H1-B lottery, they lose jobs they were already performing successfully through their practical training.

So once we cut the outsourcers away, it's not just that your candidates winning the H1 lottery are necessarily better than the American, but that they really are involuntary attrition for your company, when you are throwing away months of onboarding, and losing corporate knowledge.

So of course losing those people is terrible: Not only they were already guaranteed to be a good fit for your company's needs, but you are also losing your new hire costs, which are often north of 4 months of salary, and then you have to go on another candidate search again. 300 hours of management and engineer time for each actual new hire are not unheard of in tech, and it can go higher.

This weakens the case of the American worker being worse, but for me it shows that, if a government really is going to limit high tech immigration (and they shouldn't) , they would help US companies more if said cuts were in providing practical training. If not, then auto-approve all F1 to H1 conversions that pass muster, the way it used to be in the early 2000s, and only then do a lottery among the 'not already working here' candidates.

"It's also important to note that there are two different blocks of H1-B candidates"

This is a very important distinction.

Opening centers in Poland, Hungary & Belarus is really hot right now. Lots of well educated workers - inflation rates are crazy though.

Get rid of the corrupt, exploitative H1-B program and, in its place, introduce a Canadian-style points system so that the talented can come and stay indefinitely and not be bound to a sponsoring employer.

If only we could get the people who advocate that the most to actually agree to it!

It's probably a lost cause, at least for the near future. Under Trump, we're not likely to get any kind of relaxed immigration. Under a Democrat, we'd never get a points system. And employers of H1-Bs don't want to get rid of the program -- they like low-cost, captive employees and don't want green-card holders who could job-hop and demand market-rate wages.

H1B is one of the work visas that permits you to apply for permanent residency.

Your “low cost” comment suggests to me you’re thinking of the stereotypical Indian worker being brought in by a huge IT support outfit, not the startups being examined in this study. For these startups, especially Bay Area startups, hiring through H1B is actually a real hassle and quite expensive.

Speaking as someone who went through the H1B to green card transition, I can assure you I was anything but inexpensive :)

And, even in the case of exploitative outfits that seek low-cost workers, my experience is that the massive labor pool they draw from means quality is surprisingly high. They contract these staff out to work side by side with “local” staff and deliver work to “local” standards, and by and large succeed. The advantage they have over startups is that they are experts on the immigration process and aren’t attempting to settle workers directly into the Bay Area.

why is it corrupt?

This exactly. The present system is indentured servitude.

The American capitalist class, with the aid of its shilling apparatus (press, Congress, think tanks, academics such as the OP), wants a global labor market for the USA, in disregard of any concern other than cost savings. This fact is observable across decades of off-shoring, open borders, human trafficking to serve agribiz, lobbying Congress for more amnesties/visas/green cards, and other anti-American behavior. To aid this process along with as little pushback from the little people as possible, they've devised and promoted the ideology of "diversity" that is reputed to be "our greatest strength", making sure that this message is broadcast in a deafening din from every official organ of the shill-ocracy so that no American can see through the ideological fog and ascertain the clear fact of his dispossession brought on by a treasonous ruling class. Change my mind.

Comment of the Day from Lee!

Let's find out.

Put the H1-B's up for auction (start with half so we can retain any potential benefit from lottery options). If foreign talent is really so valuable, let's split the gains between the treasury and the firms.


Put a price on it. Immigration debate is a battle between the zero pricers and infinite pricers. The truth is somewhere in the middle and we should be maximizing the value we get out the transaction.

I also note that if we were really so starved for talent real wages would be rising fast.

It would take a lot of enforcement. Otherwise overseas parents just pay shell companies to "need" their kid.

I don't see a problem with that. If you want to pay $1,000,000 for a visa, that sounds great. You're basically prepaying the cost of your government.

Right but you've displaced the real talent which is supposed to be a public good.

That's just an argument for setting a high price or implementing price discrimination. You need to set the price at a level that compensates for the harm done. You need to pay for the government the immigrant gets, and you need to make enough "profit" to make up for the fact that you've displaced an American that would have had the job otherwise.

I do think there's an good argument for price discrimination. The 60 year-old manual laborer who doesn't speak English should be charged a lot more than the 30 year old fluent speaker with a PhD in Chemical Engineering. It might even make sense for the price to be negative in the second case.

But that's an extra layer of complexity. Just auctioning the visas would be a massive improvement.

1. Why? These are not green cards, let alone citizenship. Buying one without actual skills is pretty poor return on investment. Besides if you want to buy access for the kid EB-5 is likely a much better option.

2. Exactly what assurances do we have now that these visas actually serve the public good?

Imagine if some niche area was colluding and lowering the wages of something like compiler debugging. They all put in for H1-B and the net results is that they increase the supply pool enough to keep wages down. Even if they are in error and H1-Bs are not efficient for doing that, we would still be handing out visas without a public good.

Further, the whole point of this is that somehow startups that win the lottery gain value. Fine. Let's prove that it is something more than just having lower costs. Let's split the rewards between the public and the firms. Let's believe that markets work and see how things shake out.

Agreed. Let's do it. When is the bill going to be introduced in Congress?

Right now, the argument is between open borders and 80% open borders with a random component. Rational immigration policy isn't even on the table.

Our current immigration policy is like 5% open borders, so I really have no idea why you would think that. The reason we have an illegal immigration problem is because it's effectively impossible for 95% of the earth's population to get a legal visa to live and work in the US.

"We have immigration controls that slightly inconvenience college grads, and we have illegal immigration which we ignore." That's de facto open borders.

I smell small sample bias.

"....the results show that access to skilled foreign workers has a strong positive effect on firm-level measures of success."

If workers cannot be imported, capital can be easily exported and hire those workers at home ;)

Perhaps capital can be easily exported, but can good institutions, clustering effects, business-friendly regulations, etc? Definitely not easily to China and India, the two biggest sources of high-end talent.

Even just time-zone differences are a real problem for collaborating closely with colleagues based in India and China.

On the other hand, work gets done while everyone sleeps. Like the shoemaking elves.

Yes, and very often that work is quite useless, as for example I discovered when I tried to run code from an Indian outsourced team that actually failed to compile.

I've been through it. There is definitely a process in which you find the engineers and leads you can trust to deliver. It can be bumpy until you do.

I managed two software engineering teams - one in North America the other in Bangalore. The NA team did 4 times the work of the same sized Bangalore team. While I slept the Bangalore team would "fix" high risk bugs and break the build. The next day the NA team would clean up the mess. Because there were so many jobs in Bangalore turnover was high.

People underestimate the value of the North American (including Canada) workforce.

This was a popular idea 20 years ago, but it hasn't really worked and it's not the time zones. Getting any real productivity out of India is hard and China is going to walk off with your IP. I've seen success outsourcing to Europe, but you aren't getting a cheaper alternative that way.

It's a bit weird, but industry clusters centered around knowledge and innovation, like Silicon Valley and NYC, have survived even in the face of terrible regulatory environments, but manufacturing has proven more mobile. The right regulatory environment will draw manufacturing.

It's not about getting cheap workers, if that was the motivation the Google Zürich office with 2000 employees would be a total failure. Sometime people forget that not all the highly skilled people want to live and work in the US.

Also look at Google India headcount, "getting any real productivity out of India is hard" but that has not stopped the capital flow and growth over there. Developer anecdotes are fun to read, but the CEO and the board decisions are what matter.

H-1B visa is the lure to foreign nationals applying to US undergraduate and graduate schools. This also explains why some schools are getting even more restrictive on US nationals, even though they get a non-profit status to educate US citizens.

Pay the tuition and become a citizen or employed in the US. Colleges love it, and now, with the Chinese drop-off, are again focusing on US students.

Almost as good as getting a visa for building a building in the US.

Before comments point this out, I realize that a student visa is not the same as a H-1B, but it leads to one.

Less than it once did:

Good article. Thanks for the link.

This is probably the worst thing you have ever posted, Tyler. And that is really saying something.

Are you really sheltered enough to think that the Visa gets you access to a talent that you couldn't otherwise get?

What it gets you is the same talent you could get domestically, but at one-fourth the price.

So, ladies and gentlemen, here we have Tyler begging for US companies to be able to shut out US citizens and hire foreign workers, just to improve their profit margins. I know you hate Trump, but why take your rage out on US workers?

IPA, it is sad what a crap reader you are. I didn't even write this piece.

Legally, you cannot pay H1-B workers less than the prevailing wage. Part of the process involve a Labor Department certification that the job pays the prevailing wage in the industry.


I've never worked anywhere where H1Bs were used that way. What they are used for is increasing the size of the applicant pool and lowering the bargaining power of the applicants and existing employees. Nothing blatantly illegal, just supply and demand.

As if the prevailing wage were unaffected by the size of the labor pool.

As if a measly 85,000 visas per year makes a frigging difference.

Doug, below, points out there are 45 thousand STEM PhDs joining the market each year. I bet 85 thousand extra competitors matters to them.


How many enforcement actions can you name? Because if it isn't enforced, it doesn't matter what the law says.

The H1-B program has always been easy to cheat, because its requirements were lightly-to-not-at-all enforced. Just look at the shift from when Trump made noises about actual enforcement (not that they were really followed up on).

I've know a lot of H1-B (and other visa) holders, as I work in high-tech. Some were good. A lot weren't. Almost none of them were paid prevailing wages, even if you ignore hours because they were on salary. A lot of them had degrees from universities that, oddly enough, didn't seem to have any online presence (no matter what language you were searching in) or scholarly publications. Many of them were filling jobs that had ludicrously specific requirements that just happened to match their resume (we'll ignore the time we had to lay off almost a third of QA due to the recruiters falsifying resumes on their H1-B candidates).

No, the system has been designed to be easily corrupted, and then the few checks were rarely imposed. I'm all in favor of the auction method; there are cases where H1-B is a good thing, so let's try to keep the system focused on what it's ostensibly about. Or at least give up the hypocrisy and admit what the system was really meant to do.

A nation with tight job markets is a happy country.

The US should shoot for even lower rates of unemployment and higher job-opening to job-hunter ratios.

Immigration and imports and tight job markets...not sure the two mix....

"A nation with tight job markets is a happy country."

It's also a country that will invest in productivity growth.

I am sure some startups benefit from snagging talented engineers and scientists via the H1-B program. There may be some confounding variables though that undermine the statistics in the paper. Perhaps startups that hire H1-B holders have other characteristics that portend future success that has nothing to do with the H1-B holders’ contributions.

And there are detrimental effects to the program that are left unmentioned. The vast majority of H1-B visa winners are hired by large corporations: Ernst & Young, Apple, Amazon, etc. for entry and mid level jobs. The main effect is to lower wages for US trained accountants and engineers to the benefit of EPS compensated executives and shareholders.

The program also leads to a drain of talent from relatively impoverished countries- primarily India and China with detrimental effects on their welfare.

The program also distorts signaling that would lead to more and better US based training of future scientists and engineers. Why invest in STEM education when you can import it; at least for the time being- who knows what the future brings. Perhaps after our STEM training has been decimated circumstances will change- war, trade disputes, etc.. and we can neither import it train engineers or scientists.

There are stakeholders apart from the H1-B holders and the executives and shareholders that benefit from them. Ignore them in the name of efficiency and “absolute total welfare gains” and you get more Donald Trumps.

This entire argument is typical "fixed pie" kind of thinking. Why is there a trade-off between STEM education for immigrant students and domestic? STEM jobs aren't any more fixed in number than any other kind of job. There is probably even more opportunity for growth (e..g Google). Besides, H-1B visas have been happening for years, and it hasn't "decimated" our STEM training yet. Actually, a lot of these immigrants stay and become American citizens.

In recent years more foreign graduates of U.S. colleges opt to stay in the U.S. under the Optional Practical Training program than foreigners receiving H1-B visas.

From the link: "By the end of the 2004-2016 period, there were a total of 1,474,000 OPT approvals and 1,473,000 initial H-1B visa approvals. While both programs give foreign workers temporary employment authorization in the U.S., they are different in a number of ways. For instance, only foreign students on an F-1 visa with a higher education degree from a U.S. college or university are eligible for the OPT program, whereas any foreign worker with a degree that is equivalent to a U.S. bachelor’s degree or higher is permitted to apply for the H-1B visa. Also, unlike the H-1B visa program, which imposes an annual cap of 65,000 visas to private companies sponsoring foreign workers, there is no cap on the number of approvals available under the OPT program; all F-1 visa holders are eligible to apply. Furthermore, foreign students do not require employer sponsorship to apply for OPT, while the H-1B visa program requires employers to directly sponsor the foreign workers they intend to hire."

Phrases like "Starved for talent," "jobs Americans won't do" are the lowest sort of corporate propaganda. I thought better of you. They just want cheap labor they can control. Their desire for "talent" brought us disasters like the Boeing 737 MAX.

You mean it was these visa holders who lobbied the FAA to not require simulation training for pilots? These entry level workers were the ones making fun of the FAA? They made decisions not to check and recheck code? Help up understand please.

Americans will not accept open borders, as the short term disruption would be significant. But it would be far better to use speed bumps rather than walls. If there is high demand, let's charge for the privilege of moving to and working in the US. Charge immigrants 10-20% of their earnings. Don't let them collect most social services for the first 5 years. If they fail to earn above a threshold, kick them out after 2-3 years. That allows companies like Google or Microsoft to hire the people they want, but with a significant bias pushing them to hire American workers.

This does not obviate the need for security screening; security risks can and should still be excluded. But quotas and lotteries will never be efficient relative to putting a price on a work visa. Let's use markets where we can to make the economy more fair and efficient; that includes immigration.

I could see this effect working in two directions. The obvious is that H1-B workers bring unique skills that help young companies succeed. The second is that H1-B workers are cheaper to hire than US workers and are less likely to leave than other workers. The cheaper source of labor could be a positive when firms are concerned about burning cash.

That being said, there's probably a lot going on, and I haven't read the paper.

What unique skills do these workers bring?

Citizen workers can change jobs or demand higher pay when a better opportunity comes along. H-1B visa workers can't do this as easily. Winning employers not only get the foreign talent, they get it cheap.

This is the right answer.

Imagine the same study, but where it showed that MLB teams with high draft picks were more successful than similar teams without high draft picks. It lets the firm/team gather the surplus from the difference between what the talent could get on the open market and what they are required by law/rule to pay.

I’ve hired a number of foreign nationals under the H1B programs. All were MS or PhD from US universities, about 90% of those from “good” programs, e.g. Stanford, UT Austin. In every case, they were hired with the hope they would become career employees, were treated and paid the same as US hires, and were put into the green card process (paid for by the company) within 1 year.

About 10% were exceptional, pretty much the same as for US national hires. Most went on to moderately successful careers.

But the H1B process has changed. Today MS and the outsourcing contractors are looking for thousands of primarily entry level people. This is not start-up make or break talent. This is cheap, short term, moderately skilled and talented labor. This depresses wages and makes these careers less attractive to US citizens.

We already have more than enough H1B slots to import the best and the brightest. As several people suggest above, there are several ways to filter the flow to get the few thousand a year who are actually the cream of the crop, while not undercutting the domestic market. It’s foolish not to do this.

I've never worked anywhere where H1-Bs were systematically underpaid. Everywhere I've worked H1-Bs were used to double the size of the applicant pool and reduce the competitive pressure of needing to outbid rivals.

How does that even work? you don't apply to sponsor someone for an H1-B until after you've decided to hire them. And H1-B visa holders can't change jobs, so they are not gonna be available to interview with other companies on the off chance they might get hired. They would have to go through the whole visa lottery again.

Jesus people, get your story straight. Either H1-B visa holders are slaves OR they are gonna enlarge the labor pool and depress wages. not both. Slaves aren't free to enter the job market!

I'm not the one saying they're slaves. I never have been. They can change jobs, they just need to do so to another company of scale so they can sponsor the visa. So they are limited in their market power, and you can be harder on them, but they aren't trapped like some people are saying.

The way it works is you get 200 applicants instead of 100 for your 20 openings. If you have 200 applicants, you can offer $X and fill the positions, particularly because 100 of those are kind of desperate and you're bundling the right to live in the US with your job offer. If you have 100 American applicants, you need to offer $X*1.1 to fill the positions. You need a higher yield from the better applicants, and you get that by paying for it.

--but to put this discussion on the level of pedestrian Americans:

America already has talent: the boobtube has said so, it's had a feature dedicated to this proposition that's been popular for years, and because pedestrian Americans have been trained for even longer to be little else than naïve literalists, pedestrian Americans think to themselves without provocation: "America's got talent" . . . which is always and ever elsewhere on display, just look at the rest of our commercial "culture" as portrayed in by our corrupt and corrupting Media Establishment.

You cognitive elites have your work waiting for you.

I think the H1-B visa is a bad idea whose time has passed.

Alex could be replaced by a junior faculty member and 2 graduate students. The sum of their salaries would be about equivalent to Alex's salary + research budget. The resulting research would be fresher and less predictable. The resulting teaching would be more energetic.

Why do Alex and his ilk that they somehow deserve protection from market forces that everyone else confronts??

"Which illustrates the dirty little secret of market capitalism: Few people actually want to live under it."
If they have a better option anyway.

I doubt this. In my experience, academics is much more of a superstar market. This would be like saying you can replace LeBron James by a first round draft pick and 2 second rounders. Alex is almost certainly uncommonly gifted in doing his job in ways that aren't obvious to people not in the field that make him extraordinarily productive. Or for another analogy, think WAR in baseball. Your best argument would be to abandon tenure AND increase the number of faculty. Then you don't have to worry about WAR, because someone can just hire more people.

That's fine. If Alex is a superstar then his University will keep him employed even without formal tenure. But Alex should have to sweat to stay employed like everyone else.

Fair enough. I just think tenure is overrated as a mechanism. In my experience it is actually quite weak. How fast would Alex get fired if he said something racist that caused an uproar on Twitter? I think the bigger issue is the *organization* of a university, which doesn't allow for the same type of hiring and firing mechanisms present in businesses. Now, one could reasonably argue that the existence of tenure drives the organization structure of the university, but I think that changes the argument quite a bit. I think the implication of your line of argument is that someone like Alex can't be fired because he isn't directly accountable, and that fact could be easily changed by abandoning tenure. However, I think if GMU abandoned formal tenure tomorrow, not much would change. And it's not obvious to me that you want to change the structure of academia to be more like a business, even if one did abandon tenure.

Easy. Just make a deal to have Canada's point system and strict enforcement against illegals in exchange for half a million more legals with high skills. Bet neither the Dems nor Alex agree. It would improve the USA by eliminating millions of lower end aliens who duck up resources and increase average quality of immigrants

What's surprising about this result? Corporations benefit from doing something they constantly lobby for?

I'm a worker, not a great holder of capital. All my friends are workers. We studied math, compsci, or physics to get jobs working on computers for corporations. Corporations build teams that are 60%, 80% foreign, many of them on H1-B or having got here initial on an H1-B.

I don't particularly care if firms save a few million dollars by not engaging in bidding wars for middle-of-the-distribution domestic STEM talent. I am that talent, and I want to be fought over. I don't care if this savings makes firms incrementally more likely to reach the heights of success in a world of falling labor shares. There will still be a rank ordering in a world of a rising labor share. We workers don't view the country as some grand strategy game with assets to be moved about and deployed. We care about making money in trade the sacrifices we've made to build our skill sets.

Under the current model, the best-paying firms are stacked with foreigners and the Americans are expected to be the glue that holds it all together. It's exhausting. We're the ones who must adapt to broken, non-functional English, to rude habits, to sloppy work and cargo-cult analytical approaches that miss the forest for the trees. There's a reason these people's countries are poor and chaotic, and teaching old dogs new tricks for the sake of "the economy" isn't something we're keen on.

Your typical 2nd and 3rd world H1-Bs are not (usually) good workers or thinkers. Their role is principally to act as auxiliaries to undermine the bargaining position of the Americans who still provide most of the creative input and top-level value add. Screw the corporations and their success, bring the labor share back to the 1970s by fostering bidding wars for talent. If you want to give H1-Bs to graduates from the very best schools (top 20 say), go ahead. But these diploma mills at mediocre schools are not generating rare talent, merely cheap labor that hurts Americans.

It sounds like someone who has trouble competing. In the STEM areas, quality floats. However, being in the right place at the right time with the right skill set also makes a difference.

In the STEM areas, you have to keep current and increase you breadth as the knowledge base expands and those who don't expand their capabilities will lose out in the long term. I have spent an hour + per day all my life satisfying my curiosity.

Sounds like someone doesn’t care about American labor.

It’s all about the labor share esse. I want a high labor share, I want companies bending over backward to get American talent. You want to spam the system with foreign workers, undercutting the great mass of citizens. Pro worker? Or pro oligarch?

The thing I think you are missing is that America is not the only country producing the sorts of goods that H1-B workers contribute too. If US companies are unable to expand rapidly, then some other country will seize market share. You're basically acting just like the UAW: higher wages for me now, at the cost of making US-manufactured cars more expensive, so we lose market share to Japanese cars in the long run. Driving up wages for labor is not a thing that happens without downstream effects on American manufacturers global competitiveness, and, in the end, on wages. I.e. How much does the newer generation of autoworkers makes compared to the old? Think long term about the US tech industry and what it will look like in 40 years if we cease to be the technological leader of the world.

The above person is a different Bill. Try a different name to avoid confusion.

Thought so. The haiku was awful.

Your typical 2nd and 3rd world H1-Bs are not (usually) good workers or thinkers. Their role is principally to act as auxiliaries to undermine the bargaining position of the Americans who still provide most of the creative input and top-level value add.

If they suck so hard, how are they a threat to your wages?


The only plausible way for the american public to accept an increase like this is to reduce - or create the perception of (read: a wall) - low skilled immigration to zero. And yet, this won't be something you will see academics or public intellectuals say.

I worked for a firm importing programmers from India in the 90s. The founders started a company out of their home, IPOed in the late 90s, and eventually sold the firm to a large multinational consulting firm for a few billion dollars around 2010. We placed programmers on H1B visas that would relocate to remote areas of the country that couldn’t attract local talent. The duration of projects was typically 3 to 12 months so Americans wouldn’t relocate for such short durations.

Access to this highly skilled, hard working labor definitely made these businesses more productive by accelerating their IT project lifecycle.

The H1B visa provides opportunities for small firms to gain market share. The larger Indian firms with multimillion dollar outsourcing contracts can leverage L1 visas based on the nature of their contracts which gives them a huge advantage over their smaller competitors.

The duration of projects was typically 3 to 12 months so Americans wouldn’t relocate for such short durations.

Oh sure they would have -- at the right wage. And if the H1Bs had not been de facto indentured servants, they might not have wanted to go work in the boonies either (or would have demanded a wage premium). As far as I'm aware, oil-drilling companies are not big users of H1B visas yet manage to get large numbers of native workers to relocate temporarily to remote, undesirable areas. They just have to pay people enough to make it worthwhile.

Prevailing wage is set by the Department of Labor. The consultants weren’t indentured servants. Once they were in the country they could be poached by other consulting firms on a visa transfer to wages were subject to the same forces as native Americans. Consultants that chose to go through the Green Card process were captive for the application process, typically 24 months but once they had the Green Card they were obviously total free agents. They used the sponsoring firm to get permanent residency so there was mutual benefit.

I live in a Marcellus Shale area where the well workers typically work 12 hour shifts, 7 days per week and then return to their home for a week off. The work isn’t typical 8 to 5 office hours. Programming requires collaboration with business analysts and users in an office environment. It isn’t analogous to manual energy industry work.

New Rule : No economist can publish or promote a paper that has operational implications without receipts showing they have at least 5% of their personal worth invested in something that depends on those

As Fat Tony would say : Skin in the game or STFU

Also "translating" 1.5% diff to 23% diff is something used car salesmen and shenzen industrialists would call shady

bonus points for not giving N in the publicly available abstract

If the data were available, a nice falsification test would be to see what happened to their firm in the country the high-skilled immigrant came from. Presumably, they do more poorly when they lose their best talent.

Even more, you can begin to determine migration is a net positive. Is this a mere transfer from the originating country/firm to the US? Or is the gain to the US firm greater than the loss to the home country/firm? If the latter, the high-skilled imigrant is indeed more productive with (presumably) higher-quality complements.

The United States is Starved for Talent

We have a working population of 156 million and produce $21 tn in goods and services every year. Everyone here has to ask themselves whether Mercatus is such a bubble that no one ever challenges these howlers or whether you're trolling your readers for kicks.

Irrelevant numbers. Maybe start with 44,500 STEM PhD's awarded in 2015 or something at least close to the subject ( It would help to know something useful before lecturing others about ignorance.

Art has never been strong on self-awareness.

H1Bs have always been a joke. In the early ‘90s I spent a few years in Chicago with one of those. I was working for a management consulting firm in Milan, they sent me there to polish my skills and to get to know a larger part of the organization. I do not think they had to do any lottery, but they definitively had to spend a few tens of thousand dollars recruiting HR lawyers to demonstrate that there was no American people available to do my job. Which of course was a colossal lie. But, just as I went from Milan to Chicago, an American went to Japan, a Japanese to Germany, a German to Australia, an Australian to Milan. Only a moronic bureaucrat could find it a problem.

Working for a Mid size company who isn’t competitive for truly top, top talent I see tons of job applicants for roles requiring stem from those needing H1B. Most are recent graduates who are applying near or after their graduation date meaning other companies have passed on them. They do horribly on our behavioral and technical interviews as a group. It makes me wonder if we really do need more H1Bs and if it isn’t a quality be quantity issue.

Why not just have wages rise, and then companies can find the talent that they seek.
It's amazing how people want to take a successful America, with a prosperous middle class, and crush their wages with mass immigration.
Boost your per capita GDP, that is what matters more than a companies profits or aggregate GDP

The education racket is too expensive and people don't really need 4 year degrees. I don't think many managers realize this. They want only country club members.

And prospective candidates are discouraged by this system.

"Put differently on some margins the US is starved for talent."

Only because creating talent costs money.

The high cost of creating talent: free education, trade school and apprenticeship programs, etc, destroy wealth. The destruction of wealth kills creating talent.

Just the same as the high cost of paying workers destroys wealth, and wealth destruction kills jobs.

Wealth is based on profit, (money not paid to workers, or to create talented workers), so only by not paying workers or not creating talent can the wealth be created to demand talented slave imports.

I grew up in Indiana when the State Merit Scholarship of $1000 per year was enough to pay the subsidized tuition at IU, Purdue, Ball State, but I envied California where Reagan was railing against the leftist universities and colleges with zero tuition for California born, and out of State tuition was the same as Indiana resident tuition at Indiana State schools.

At the time, Purdue had equivalent status in both engineering and ag as its California peers, both below eastern schools like MIT, CMU, and others. This was before "computer science" and computers were a product of electrical engineering, plus some math dept help, with women working as calculators and computers.

Schools in the east were more stovepiped due to the funding, while in California, "free education" allowed more a la carte so more interdisciplinary work was possible creating a new field of both study and work. Students were not required to limit classes, advisors, to meet scholarship requirements.

This is why the center of gravity in technology moved from the east around Boston but out to Syracuse, CMU, Purdue, to California. In my opinion.

Immigrants have generally cut across stovepipes, mixing and matching education and work across disciplines, often finding people to help them along the way.

As the influence of Reagan and those who backed him, the stovepipe of profit focus kills talent creation.

Remember, profit is the money not paid to workers or to create talent.
Immigrants are free of

Tyler: The size of the effect should have you scrutinizing the methods, not advertising the results.

> If anything "it's harder to get a job as an H-1B" means they'll be willing to settle for less from employers, and that in turn drives down the wages of Americans competing with them.

That seems like the comments from the non-skilled domains. On the contrary, it is the large local "under-employment rates" (local graduates working in jobs that do not require any university degrees) that drive down the salaries.

From the NSF report, Business Domain Professional Salary (excluding University Domanin):

BA Graduates:
NativeBorn | TempResid | Sect
61000 | 81000 | BA
80000 | 84000 | SciEng
56000 | - | BioMed
80000 | 85000 | CompMath
81000 | 85000 | CompInfo
59000 | - | Phy
60000 | - | Chem
48000 | - | SocPsy
85000 | 80000 | Eng
109000| 83000 | SciEngMgr
70000 | 77000 | SciEngTech
68000 | 116000| Mgt

Engineers normally require local licensing without which they might not be able to sign off projects. At PhD level the better H1Bers will normally have naturalized, thus the TempResid sample is biased.

PhD Graduates:
NativeBorn | Naturalized | TempResid | Sect
102000 | 125000 | 93000 | PhD
106000 | 119000 | 90000 | SciEng
103000 | 106000 | 51000 | BioMed
115000 | 117000 |106000 | ComMath
114000 | 117000 |108000 | ComInfo
115000 | 110000 | 75000 | Phy
111000 | 110000 | 74000 | Chem
121000 | 130000 | 90000 | Eng
134000 | 156000 | - | SciEngMgr
95000 | 131000 | 87000 | SciEngTech
102000 | 131000 |147000 | Mgt

There are enough IPOs happening to get statistically reliable data for this kind of study?!

I would bet that when Alex closed with "if we play our cards right the world's best could be the U.S.'s best" he meant we should open a path to US citizenship for talented people, and is now chuckling about the direction these comments have taken.

Comments for this post are closed