The H-1B/Age Connection: So Simple to Explain, So Hard to Understand

If you are a serious follower of the Great H-1B Debate, you may have seen the recent pair of dueling op-ed pieces, one by Ross Eisenbrey  of EPI and the other by Alex Nowrasteh  of the libertarian Cato Institute, concerning the H-1B work visa. Eisenbrey has generally been skeptical about the visa, while Nowrasteh is an outspoken supporter of both an expansive H-1B policy and of liberalized immigration policies in general.

I have points of disagreement with both of the essays, and as you might guess, find most of Nowrasteh’s piece to be based on faulty premises and misinformation. However, there is one kernel of wisdom in his article (emphasis added):

Migrants with H-1Bs are typically young, highly educated and earn high wages. In 2014, 94 percent of new H-1Bs were under the age of 40, while 99.8 percent had at least a bachelor’s degree and 54 percent at least a master’s degree.

This really amounts to a Freudian slip by the pro-industry Nowrasteh. Yes, indeed, this is one of the two primary appealing aspects of H-1Bs for employers — they are young. And young means cheap. Even with a (hypothetical) employer who might give wage parity to his young foreign and domestic workers, that employer is saving a bundle, in both wages and benefits, by hiring young H-1Bs in lieu of older (35+)  Americans.

Though there are other factors, notably a desire for immobile labor, for many employers, H-1B is fundamentally about age. Unfortunately, it’s very difficult to get even those who are critical about H-1B to keep this simple fact foremost in mind.

Take for instance the much ballyhooed recent incident in which the giant utility Southern California Edison laid off American IT workers and replaced them by much cheaper imported foreign workers. People were properly outraged, but in all the discussion I haven’t seen anyone seriously raise the question of WHY the foreign workers have such low wages. Yes, some people have correctly mentioned loopholes in the legally-required prevailing wage, fine, but the overwhelming reason SCE can get away with paying those foreign workers much less than the Americans they replaced is that the foreign workers are much younger than the Americans.

Just as Nowrasteh let the cat out of the bag on the age issue, “planted” articles in the press promoting H-1B often unwittingly do the same. An excellent case in point brought to my attention this evening is an April 3 Houston Chronicle piece, “Scramble Is On for Coveted Work Visas,” by Lomi Kriel, containing the priceless passage:

But though [medical device startup Cognita CEO] Gaurav Patel] cast a wide net, 80 percent of the eligible pool turned out to be foreigners. The only American applicants had decades of experience, not a good fit for the entry-level position. So Patel hired a 24-year-old from Mumbai who had just graduated with a master’s degree in engineering from the University of North Carolina at Charlotte, where he specialized in robotics.

And there you have it, a perfect example of Americans being shunned in favor of a foreign worker (and by the way, NOT involving an IT services outsourcing firm). The fact that Patel considers the Americans overqualified should be Patel’s problem, not something for the U.S. Congress to solve for him. On the contrary, Congress should be wondering how in the world they ever wrote laws that allow such a thing.

This is, as I said, one of the core elements of H-1B, which I see all the time, including coincidentally this afternoon, when a well-qualified American applicant told me he had been rejected from a certain up-and-coming employer because he was, in the words of the phone interviewer, “overqualified” for the job in question. Since this American has also applied for other positions in the company without even a phone interview, he is presumably underqualified for those jobs. So my friend is both underqualified and overqualified to work at that firm — doesn’t leave much, does it? And yet I happen to know that this firm does hire foreign workers, who somehow have “just the right amount” of qualification.

What a sad but telling example of how, as pointed out by Neil Ruiz of Brookings from a different point of view, “Our immigration system is broken.”

Interestingly, this would be embarrassingly easy to fix:  Abolish the four-tier experience levels system used in determining prevailing wage, i.e. remove the financial incentive to give hiring preference to young foreign workers over older Americans. Define the prevailing wage to be the 50th (or better, the 75th) percentile for the given occupation in the given region — WITHOUT breaking it down according to experience levels. No solution is perfect, and there would still be room for employers like Patel to play games, but this simple solution would be a big step in the right direction.

But NO useful steps will be taken if the age issue is not on the radar screen of even the critics of H-1B, in spite of being so central to the problem. I urge those of you readers involved in proposing policy to keep in mind HOW those foreign replacements at SCE could be so cheap; the answer is a three-letter word.


12 thoughts on “The H-1B/Age Connection: So Simple to Explain, So Hard to Understand

  1. I can remember a time when you had to be over 60 to be designated as an “Old F@rt”. Its interesting that in the IT world, you get that designation by being over 35. Unfortunately, I can attest to the fact that 35 is the real number where it starts getting harder and harder to get work in software. I’m over 50 now and haven’t seen an IT job in 5 years. I had an interview a year ago, and now I’m considered stale because I haven’t worked in so long. I still write code and build circuits, but these days its just for fun. I mean, who wants to mow the grass when you can build a robot to do it for you. However, playing devil’s advocate, I would be lying if I said that I am as sharp as I was 30 years ago. Yet, I can still code faster and better than some of the younger guys. As far as what to do about companies that hire H1-B’s over Americans, the simplest thing I can think of is to tax the companies who hire H1-B’s. If they had to pay a $100K /year tax on every H1-B they hire, then they might consider more Americans for the job. But sadly, I don’t think that or any other solution will ever see the light of day as long as the corporate lobbyists are in charge.


    • I’d already been using the 35 cutoff for years when the 2011 GAO report came out, with an interesting graph that showed the same thing. It showed the age distribution of H-1Bs and Americans, and the two curves crossed right at 35. 😦

      I actually believe I’m a better programmer today than I was 20 or 30 years ago. I can learn new languages faster, fix bugs more quickly and better avoid them in the first place. Experience does count.


    • I forgot to mention that I still see a lot of job listings saying they want *exactly* 3-5 years experience. While a lot of people would assume that meant a minimum of 3-5 years, what I have seen is that it does not. I know another guy who deliberately left the dates off the college experience on his resume and truncated his job history to the past 4 years specifically for that reason. His age was cleverly concealed on his resume. For some reason, he always got jobs too – and he is in his 60’s. It was only when they demanded an in-person interview instead of a phone screen, where they could see his gray hair, that he had problems.


  2. If there is one thing that gripes me about STEM employment it is the age issue, but more or less separately from H-1B. When I started working in the mid-1970s, the rule of thumb was that a newbie got $X, if you were good and progressed in five years you got $X*2, and if you were good and lucky, in twenty years you’d get $X*4. There seemed to be some instances of all of these on the staffs I had direct experience with. Well, fast-forward 40 years and now, with many quibbles about the rate of inflation, a newbie gets $X*1/2, in five years you might get $X*2/3, and in twenty years you get $X*3/4, unless you’ve progressed to management and a vanishingly small number ever do that, topic for another day.

    These numbers should preclude the bogus idea that the compression comes from having raised newbie rates – has not happened, quite the opposite. And salaries effectively top out, at much lower numbers, at about 5 years experience. IOW the industry, as far as I can see, has eliminated the senior experience tier, and seems to be pretty much STEM-wide, and barely pays for even the lesser levels of experience.

    I guess this cuts against Norm’s age thesis somewhat. Not that the age thesis isn’t still true, but the leverage rather smaller, at least, and it may be a bit of correlation rather than causality.

    Now, one may (must) ask, *why* has the industry done this, if not for money? I want to offer two ideas. First, that especially in the matured, commoditized field, business anthropology prefers to employ mediocre talent, NOT top talent. Experienced staff would be a threat to the phony-baloney managers who make the only good salaries in the building. They want to get rid of people before they reach that experience level. The second argument is that this is a historical artifact, like the human appendix, that for some short time in the 1960s and 1970s, as the first computer science programs were coming up at the university, it might well have paid to dump your old punched-card staff for some bright new CS graduates. And this has become ingrained in the industry. Recruiters and lawyers get rich over staff churn and H-1B. It hasn’t made sense for thirty years but still goes on for these exogenous reasons. People are weird sometimes.


    • There are really two main issues concerning the harmful effects of H-1B on American workers: Wages and job opportunities. Your comments regarding wages are of course right on the mark, but that is only part of the story. There are some older American workers who go through quite miserable periods in which they simply cannot find work in their field; many are forced to leave.


      • Is there little wonder that women and minorities choose different careers. Those just entering school have analyze the situation and determined that the long term opportunities are not in IT and many other STEM fields.


  3. To his credit, Nowrasteh did say “The H-1B should not be company-sponsored; it should be held by the worker and he or she should be able to switch jobs without government permission.” Maybe like me you find it an effort to focus on some of the pro H-1B editorials at times because of what Eisenbrey calls the ‘bunk.’

    Perhaps Nowrasteh’s view reflects the libertarian attitude of The Cato Institute, though I’ve seen liberal libertarian (?) Lawrence Lessig scoff at Cato as corporate rather than libertarian. It was founded as the ‘Charles Koch Foundation’ in 1974, and there was a kerfuffle in 2012 when some at Cato felt the Koch brothers were trying to increase their control of it and make it more purely a Republican party organization.


    • Giving the foreign worker full control over the visa would fix one of the problems I’ve cited, which is the desire among employers for immobile labor. However, it would do nothing to solve the age problem, which was the theme of my posting. Again, this is an absolutely central issue in H-1B. Unfortunately, those who propose policy are oblivious to it.


  4. Interesting piece[1] in Ars Technica yesterday (S 25 Apr 2015) on an age-discrimination lawsuit against Google, with employee-median-age data scraped from Has anyone tested “the H-1B/age connection” for a (presumably negative) correlation between H-1B hirings and median age of technical employees? (Apologies for not knowing the literature.)



  5. @tlroche April 26, 2015 at 10:28 pm: ‘Has anyone tested “the H-1B/age connection” for a (presumably negative) correlation between H-1B hirings and median age of technical employees?”

    @matloff April 27, 2015 at 6:11 am: ‘Hard for there to be a correlation when they ALL are doing it’

    Umm … either I’m missing something or you are.

    Note that I used the plural ‘hirings’ and not the singular ‘hiring’. I.e., I hypothesize a negative correlation between the quantitative variable of an organization’s H-1B hires (e.g., relative to all its technical hires over a given period, or relative to its entire technical workforce) and the median age of its technical workforce. Your point would make sense if

    1. I was proposing a correlation between a binary variable (i.e., does the organization hire H-1Bs or not) and median age of its technical workforce. But I am not; nor, frankly does that seem reasonably inferred from my comment.

    2. all organizations with technical workforces hire H-1Bs to the same extent/intensity: e.g., every organization hires the same amount of H-1Bs relative to {all its technical hires || its entire technical workforce} over a given period. ICBW, but I strongly suspect that is empirically false.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s