Once Again, Confusion on “Good” and “Bad” H-1B Employers

Sorry for my lack of blog posts recently, just too busy for it. But I must comment on two interesting Computerworld articles today, both with female protagonists. The first wonders whether Ted Cruz’s choice of Carly Fiorina to be his VP will result in a clash between the two of them regarding H-1B. The other reports that the fabulous Jennifer Wedel is still active in her speaking out against the visa.

Cruz introduced the strongest bill to date reigning in H-1B, while Fiorina, as former CEO of HP, would strongly favor the visa program. But I see no clash at all. If Cruz wins the presidency, he will do exactly what Donald Trump did — backpedal, saying that the H-1Bs hired as foreign students at U.S. universities are the “good” H-1Bs, while those imported directly from abroad, via the “rent a programmer” firms such as Infosys, are the “bad” ones. Fiorina’s old firm hires from the U.S. campuses. Hence no conflict.

The Computerworld article points out that HP also does some offshoring, but I believe that that is mostly to HP outposts abroad, rather than going through Infosyses. So, Cruz and Fiorina would announce support for a bill clamping down on the Infosyses, with business as usual for firms like HP. This would be highly misleading, but this is the pattern we’ve seen repeatedly in proposals to “reform” H-1B.

By the way, I was quite impressed by Fiorina during the earlier primary debates. As a Democrat, I disagree with her on lots of topics, maybe most, but Cruz made a good choice in her, I believe.

And Darin Wedel seems to have made an even better choice with Jennifer as his spouse. As the article reminds us, Mrs. Wedel thrilled many of us in 2012 by asking President Obama point blank why he supported H-1B when many qualified engineers such as her husband had trouble finding tech work. The look on Obama’s face when she told him that her husband was a semiconductor engineer was priceless, as was his reply, “We’re told that people like your husband should be able to find work right away.”

Judging from Obama’s expression, I’m sure that he really believes it. And why shouldn’t he? That’s what people keep saying. Texas Instruments, which also hires foreign students from American campuses, is one of the “good” employers of H-1Bs, the claim goes, and even many critics of H-1B believe it. The fact that TI laid Darin Wedel off, but hired H-1Bs (including some whose jobs he could have done), should cause some cognitive dissonance. But no, this artificial distinction between the foreign student H-1Bs and the directly imported H-1Bs is so entrenched that they just don’t see that TI is equally culpable as the Infosyses.

The Computerworld article falls into that trap too:

Although Darin Wedel didn’t train a foreign replacement and was laid off with other workers, Jennifer Wedel says Texas Instruments was lobbying for increasing visa use prior to the layoffs, and she sees a connection.

Of course there is a connection. Wedel was laid off by a company that hires H-1Bs, some of whose jobs I am sure Wedel could have done. So what if he didn’t have to train them? Wedel is just as much a victim as are those laid off by Disney, SCE etc.

Again, there is no way to straighten out the H-1B mess unless people truly understand the nature of the problem.

By the way, in 2012 Computerworld thought that the Wedel incident would become an election campaign issue, which of course it did not. It did briefly become an issue this year, thanks to Trump, but it quickly faded. It never arose once during the Democratic debates, even though Sanders and Clinton disagree on H-1B.



Further Compelling Evidence Against “Staple a Green Card”

As most readers of this blog know, I strongly oppose proposals to “staple a green card to the diplomas” of foreign STEM students earning MS or PhD degrees at U.S. universities. Those proposals are based on the premises that (a) we have a shortage of people with STEM grad degrees and (b) the foreign STEM students are Einsteins-in-waiting. I’ve shown numerous times that neither of these premises is valid.

For example, concerning (b) above, foreign students in computer science, putatively the most important STEM field, are on average WEAKER than their American peers: The foreign students file fewer patent applications; they are less likely to work in R&D; and among those with doctorates, they earn their degrees at less prestigious institutions.

Today’s edition of Inside Higher Ed shows data that is quite disturbing concerning (a).

Percent of Doctorate Recipients With Job or Postdoc Commitments, by Field of Study

Field 2004 2009 2014
All 70.0% 69.5% 61.4%
Life sciences 71.2% 66.8% 57.9%
Physical sciences 71.5% 72.1% 63.8%
Social sciences 71.3% 72.9% 68.8%
Engineering 63.6% 66.8% 57.0%
Education 74.6% 71.6% 64.6%
Humanities 63.4% 63.3% 54.3%

So, the new PhDs in engineering aren’t doing much better than those in the humanities. In fact, PhDs in the social sciences and education are doing better than those in any STEM field.

Even some immigration-reform people have endorsed “Staple” as long as it is restricted only to PhDs. The above figures show such views as ill-advised.

And getting back to (b): An Australian reader brought my attention to this article regarding the University of Sydney. The article is mainly about a sudden change of date for a graduation ceremony and resulting hardship on some students from China and their families, but this passage is interesting:

It is the second time in the past year that the university has come under criticism for its treatment of international students. Last year, it is understood the Chinese consulate appealed to the university when 37 per cent of Chinese students were failed in one of the university’s business courses, drawing accusations of bias towards non-native English speakers.

Though it does appear that the Chinese students are getting a raw deal regarding the graduation date, that 37% figure is troubling. As you may recall, I recently discussed a Wall Street Journal article titled “Heavy Recruitment of Chinese Students Sows Discord on U.S. Campuses,” which among other things cited U.S. faculty who say they are lowering standards in their courses as a result of the low quality of Chinese students.

Even more disturbing is the Chinese consulate’s involvement, protesting that the business course was unfair to the Chinese students. The reader who called my attention to this article added that a professor had told him it is common for Chinese students to berate professors who give them low grades. I recently taught a graduate class in which a Chinese student was quite rude to me, stridently claiming that the (oral) interactive grading for homework was unfair to her, due to language issues. When I told her that two of the three students in the class to whom I gave A+ grades were from China, with English no better than hers, she had to back off. But this student is both weak and has a lousy attitude; we already have enough weak American students with bad attitudes; do we need to import more, via “Staple”?

All of this is quite disconcerting. I must add that it is certainly not consistent with traditional Chinese culture; in the past, a Chinese consulate would not have even considered objecting to the 37% failure rate, out of shame. How things have changed!

As I said in my previous post, there are many truly excellent international students, including from China (over the past weekend I wrote a letter of support for the green card application of one), but it’s clear that a blanket policy like “Staple” is just plain wrong.

H-1B and Gender: Much Ado About Nothing

The April 1 Computerworld article, “How many H-1B workers are female? U.S. won’t say,” is interesting in its pointing out how touchy the federal government has become over the H-1B work visa issue. The feds’ claim that the information would be difficult to retrieve seems fishy.

But the IEEE-USA’s claim the men are overrepresented among H-1Bs compared to among Americans in similar occupations is fishy too. Quoting from the Computerworld piece:

The best source of data for lawmakers on the gender of H-1B workers has been the IEEE-USA. In 2013, Karen Panetta, a professor of electrical and computer engineering at Tufts University who was representing the IEEE-USA, testified before the Senate Judiciary Committee and told it that as many as 85% of the visa holders are men.

“Best source?” Really? Here is what Panetta (yes, daughter of the Panetta) said in her testimony:

My own experience tells me that the vast majority of H-1B workers are men. Everybody knows this. The IEEE-USA represents more American high tech workers than anybody else, so we have sources. One from inside the industry, looking at the offshoring companies that dominate the H-1B program, is that their global hiring is 70% men. But in the US, where outsourcing companies get more than half the capped H-1B visas, the ratio is more like 85% men.
“Everybody knows this?” A common way of saying, “We don’t really have serious data on this.” “We have sources”? Tell us who. And tell us if that was just the person’s “own experience,” like that Panetta refers to, versus careful data collection and analysis. And what is this “more like 85%”? This is a scientific statement, “more like”? Does Congress really invite “experts” like this to present testimony? Actually, Panetta then admits she really doesn’t know:
As an engineer, I don’t like making decisions without hard data. The IEEE-USA has been trying for months to get the actual data on this from DHS. They have been stonewalling us. It’s a simple question: how many women get H-1B visas?
Again, I am not defending stonewalling, but Computerworld’s statement that Panetta has the best data is unwarranted, and indeed the article’s own data is not bad. The H-1Bs are mainly programmers, i.e. they fall into the two government categories shown in the article’s bar chart, Programmers and Software Developers. (The government shouldn’t have two separate categories, but that’s a different issue.) And in those two categories, women are MORE represented among the “India-born,” Computerworld‘s proxy for H-1B/former H-1B, than among the general population in those occupations. IEEE-USA ought to be praising H-1B for increasing female percentage, not darkly hinting that the government is covering up some sexist cabal.
As I’ve said bluntly before, IEEE-USA has an agenda, which is to punish the “Infosyses,” thus deflecting attention away from their equally-capable allies, the “Intels.” And they have no credibility, given their refusal to back up their ever-present claim, “We represent 227,000 American engineers” by taking a poll of their members on the issue. Once again, my favorite quote from Senator Grassley: “No one should be fooled.”



Pro-H-1B Study Shows What’s Wrong with H-1B

Stuart Anderson has long written in favor of the H-1B work visa, and indeed makes his living from it, via his (apparently one-man) organization, the National Foundation for Public Policy. He also wrote the 2000 H-1B law as a Hill staffer, and later was at the USCIS. He’s getting publicity for his latest study in various places, including a Wall Street Journal blog.

Yet, oddly, the study confirms what I’ve been saying for a long time: The H-1Bs are generally NOT the best and the brightest, and the nation would benefit by changing the visa program to focus on the outstanding talents. Anderson’s findings, in other words, show exactly what is WRONG with H-1B and various proposals to reform it.


  • Anderson finds that immigrants are founders (OR cofounders, a point I’ll come back to) of 51% of the billion-dollar startups. Yet immigrants form more than 51% of Silicon Valley techies, so once again we see that the immigrants are UNDERperforming.
  • Among those immigrant founders, 14 (out of 44) are from India, 32%. Yet Indians form 70% of the H-1Bs. So the Indians are underperforming too. Most of the founders are from countries from which we have rather few H-1Bs, such as Western Europe, Canada and Israel.
  • Only 1/4 of the immigrant founders came to the U.S. as foreign students, counter to the industry’s claim that foreign students are the best of the H-1Bs.
  • Of those with U.S. degrees, most are from the really top schools, such as MIT and Stanford — quite a contrast to the Staple a Green Card proposals, which would give an automatic green card to any foreign STEM graduate student, no matter how weak the school is.

As usual in these “studies,” anyone born abroad is considered an immigrant, such as an example Anderson cites, Kenneth Lin of Credit Karma. Lin immigrated to the U.S. with his parents at age 4, hardly the poster boy for the H-1B work visa program that Anderson implies. Others on Anderson’s listed cofounded their firms with Americans, yet he gives them full credit.

Before H-1B, we had the old H-1 visa, titled “Aliens of Distinguished Merit and
Ability,” basically the “best and brightest” theme. That did deteriorate over the years, so that anyone with a college degree because “distinguished,” but it certainly had the right intent. Anderson’s study shows that H-1B doesn’t fulfill that intent at all.


U.S. Universities Souring on Students from China?

A reader called my attention to this Wall Street Journal article, titled “Heavy Recruitment of Chinese Students Sows Discord on U.S. Campuses.” My wife, a Chinese immigrant, also mentioned it to me.

The concerns described by the article are not new: Lack of mixing of the Chinese students with the Americans (more of a problem with the Chinese than with the other international students); lack of analytical skills; pressure felt by some professors to lower standards in order to accommodate; and so on. But the sheer volume of students from China today is really putting all this under the spotlight.

As many readers of this blog know, although I strongly support facilitating the immigration of the world’s “best and brightest,” I am strongly opposed to the “Staple a Green Card to Their Diplomas” proposals, which would give automatic green cards to all foreign students earning STEM Master’s or PhD degrees at U.S. universities. This WSJ article, on which I will comment below, illustrates why “Staple” is a bad idea.

As the article points out, universities, both public and private, are under great pressure to treat foreign students as cash cows, for instance because some schools charge them higher tuition than for domestic students. It should be abundantly clear that this will necessarily cause a lowering of standards, something usually whispered but here discussed openly by a major newspaper. I was told by a colleague at a mid-level university just yesterday that her Computer Science Department has 400 (!) international students in its Master’s program. Since China leads all nations in the number of foreign students in the U.S., the article is even more relevant to the “Staple” issue.

We do have some “best and brightest”-class students from China. Whenever I teach a graduate class, there is typically at least one Chinese student among the top two or three students. I actively help top foreign students find jobs in Silicon Valley. But most of my Chinese students struggle in my class, for the reason cited in the article, profound weakness in analytical skills.

This in turn is due to a lack of development of these skills in China. With a rote-memory educational system and an age-old culture that does not question authority, the sharp, insightful, innovative students will necessarily be in the minority.

But the article gets it wrong in some senses. For instance, it quotes Professor Nicol on the weakness of many Chinese students in expressing themselves, with the implication being that this is a language problem. That is true to some extent, but the much larger issue is that many of them can’t express themselves well even in Chinese. Traditional Chinese education simply does not develop that skill, and in fact arguably hinders it.

Coincidentally, I was just discussing the language issue the other day with a U.S.-native graduate student. It’s interesting that Nicol “excises colloquialism from his lectures to avoid confusing the nonnative English speakers.” I do use idioms, but explain for the non-native speakers, in a way I aim to be fun for all students, both foreign and domestic. But again, I wish to emphasize that the low analytical abilities should not be attributed to language problems.

In any case, the language deficiencies can be attributed in large part to the failure of most Chinese students to mix with the Americans. It is real (and mostly special to the Chinese among all foreign students), but I must confess that in spite of decades of closeness to Chinese communities, both academic and personal, I have no idea what causes that failure to mix. Is it due to lack of interest on the part of the Chinese students, or lack of welcoming on the part of the Americans, or both? It is interesting that Mr. Shao applied to, and was accepted by, a fraternity; I doubt that this is common, but it is an encouraging sign.

Most troubling, though, is the article’s theme that many professors feel pressure to reduce the quality of their courses, to accommodate the Chinese “customers.” I have been told that this has occurred in Australia as well.

Again, the severe lack of proper analytical abilities is not a new concept, and East Asian governments have been trying to remedy it for years, without much success. But given centuries of tradition, changing this is “easier said than done” — precisely one of the idioms I recently explained to my graduate class.




DHS Sees What It Wants to See

Much appreciation to an alert reader for pointing me to DHS’ Final regulations on OPT. Final, that is, subject to the pending lawsuit, but this is not my subject in this post. Instead, I wish to point out an outrageous mischaracterization DHS makes of my EPI research paper in that document, exemplifying the continuing grossly irresponsible behavior of that agency.

DHS says (referring to me as “EPI,” emphasis added),

With respect to the studies by Dr. Hunt, DHS notes that the NPRM cited those studies in support of the general proposition that STEM workers “are fundamental inputs in scientific innovation and technological adoption, critical drivers of productivity growth in the United States.” 80 FR 63383. The EPI study did not question this proposition. Rather, the EPI study examined a narrow band of STEM fields to show that “immigrant workers, especially those who first came to the United States as international students, are in general of no higher talent than the Americans, as measured by salary, patent filings, dissertation awards, and quality of academic program.”  (61) Specifically, the EPI finding is focused on whether foreign-born students who earned computer science and electrical engineering degrees in the United States file patent applications at higher levels than U.S.-born students earning the same degrees. For electrical engineering, the analysis showed that patenting activity of U.S. and foreign-born students was about the same, while for computer science the analysis showed that foreign-born computer science students apply for somewhat fewer patents than do their American peers.

The EPI paper, however, acknowledges that the Hunt studies cited in the NPRM cast a much broader net, encompassing a myriad of science and engineering fields. The Hunt papers considered the impact of foreign-born workers employed in the United States in myriad visa classifications and fields of study, and was not focused solely on F-1 students or STEM OPT students (nor to just Computer Science and Electrical Engineering research activity). As explained in the Hunt papers, there is support for the proposition that foreign-born scientists and engineers achieve higher rates of U.S. patent filings. The Department continues to believe such patent rates support the conclusion that the STEM OPT extension is in the national interest.

This is way, WAY off the mark, in numerous aspects.

First of all, Professor Hunt published two papers, DHS references 28 and 29. (The latter originally was first published as an NBER working paper, then later in the Journal of Labor Economics, JOLE.) Reference 28 boils down to saying that immigrants have a lot of patents, because they tend to be in STEM fields; American bank managers, truck drivers, lawyers, doctors etc. generally don’t file patent applications. But in Hunt’s second paper, Reference 29, she looked at PER-CAPITA rates; relative to Americans in a given profession, do immigrants file fewer patent applications or more of them? Her answer was that the immigrants are LESS likely to file for patents than natives. In my EPI paper, I quote Hunt from her NBER paper:

After I control for field of study…and education…both main work visa groups and student/trainee visa holders have statistically significantly lower patenting probabilities than natives.

Academic politics being what they are — especially concerning research on immigration, where reviewers often have polarized views on this matter — JOLE told Hunt to remove H-1Bs from her paper, so the above passage was not in the final JOLE version. Nevertheless, even with that omission, she still found that the immigrants filed no more patents per capita than did the Americans.

So, DHS is misrepresenting not only me but also Hunt.

It gets worse:

Recall from above that DHS says

Rather, the EPI study examined a narrow band of STEM fields to show that “immigrant workers, especially those who first came to the United States as international students, are in general of no higher talent than the Americans, as measured by salary, patent filings, dissertation awards, and quality of academic program…For electrical engineering, the analysis showed that patenting activity of U.S. and foreign-born students was about the same, while for computer science the analysis showed that foreign-born computer science students apply for somewhat fewer patents than do their American peers.”

I showed that and much more. Compared to the natives, the former foreign CS students not only filed fewer patent applications, but also had lower salaries (even after getting a green card and becoming free agents), and among those earning a doctorate, earned their degrees at lower-ranking universities. Both EE and CS immigrants were less likely to work in R&D, directly contrary to DHS’ claim of “innovation.”

And as to DHS’s saying that my study only covered the “narrow band” of computer science and electrical engineering: First of all, that’s hardly narrow! Computer occupations alone account for 50% of the H-1Bs. But even more important, DHS itself notes that Hunt’s paper covered the gamut of fields, and as I noted above, she found that “work visa groups and student/trainee visa holders have statistically significantly lower patenting probabilities than natives.” Even if DHS wants to dismiss Hunt’s NBER version of the paper, her final JOLE version (which excluded the H-1Bs) found that, again as noted above, the immigrants filed no more patents per capita than did the Americans.

In other words, DHS’ statement, “As explained in the Hunt papers, there is support for the proposition that foreign-born scientists and engineers achieve higher rates of U.S. patent filings,” is, if you will, patently false.

Furthermore, my paper also quoted the “wide band” statement of Bound et al:

In physics, biochemistry, and chemistry much of the expansion [from the mid-1980s to mid-90s] in doctorate receipt to foreign students occurs at unranked programs or those ranked outside the top 50; the growth in foreign students in engineering is distributed more evenly among programs. Among students from China, Taiwan, and South Korea growth has been particularly concentrated outside the most highly ranked institutions.

This of course jibed with my similar finding that the foreign CS students earn their degrees at lower-ranked schools. But no mention of my Bound citation by DHS in their “analysis” of my paper.

Note by the way DHS’ repeated use of the term “foreign-born” rather than simply “foreign.” As I have pointed out, this is official language from the industry lobbyists’ literature. Granted, a minor point, but one that exemplifies DHS’ cozy relationship with the lobbyists.

So, was that relationship so cozy that the DHS deliberately mischaracterized my EPI paper? Or are the DHS people such “true believers” in the glories of foreign work programs that they simply couldn’t see what was in plain sight in my paper? Maybe they simply lack critical reasoning skills? Whatever the cause is, this DHS document (I’ll discuss other parts of it in future posts), represents a gross failure in responsibility of a key government agency.


What a President Trump or a President Sanders Could Do

Both Donald Trump and Bernie Sanders have expressed a desire for genuine reform of the H-1B and related programs. Trump’s written platform is strong (though some of his statements leave some abiguities), and Sanders has spoken out against H-1B for years. Let’s fuse the two of them into a single President Trumpers, and ask what would he could actually accomplish in the foreign tech worker area.

It is a given that no matter how strongly President Trumpers would feel about the issue, Congress — doing the bidding of the powerful tech industry and its powerful allies — would fight him every step of the way. But there is actually much he could do on his own, through legitimate executive action.

The most obvious action would be rescind the Bush-Obama executive actions extending OPT, with the time reverting back to the original 12-month duration. An even bolder act would be to disband OPT entirely, though I wouldn’t recommend that unless meaningful H-1B reform were to be achieved. (I also would suggest grandfathering current OPT “students.”)

A less obvious, but very effective action would be to give the proper interpretation to the term H-1B dependent employer, which is defined to be any employer for whom at least 15% of the workforce consists of H-1Bs. The current interpretation is absurd, as it includes all of the employees in a firm, e.g. the receptionists, delivery people, janitors and so on — jobs that certainly don’t fit the H-1B statute’s definition of workers eligible for the visa, i.e. “speciality occupations for which a bachelor’s degree is normally required.” In other words, since the ostensible aim of the law’s defining an H-1B-dependent category is to protect American tech workers, the denominator on which that 15% threshhold is based should consist only of tech workers, i.e. the company’s engineering and technical workforce, not its general workforce. This would have a huge impact.

President Trumpers could also declare that the green card process will not allow employers to reject Americans on the basis of being “overqualified.” As I have often mentioned, one of the major reasons employers like H-1B and related programs is that they allow them to bypass qualified but older American applicants, hiring young foreign workers instead. This change to the green card process would be of substantial help here.

Most Americans are probably unaware of the fact that many foreign STEM students are supported by federal money in the form of being hired under the research grants held by their professors. President Trumpers could direct his major research funding agencies, such as the NSF, NIH and so on to give funding priority to professors who have a solid track record of hiring U.S. citizen and permanent-resident students and post docs.

Similarly, Trumpers could direct his funding agencies to give priority to universities that have higher-than-average proportions of domestic students in their STEM graduate programs. Needless to say, he would also threaten to reduce federal funding for universities that view foreign students as “revenue enhancement sources.” Trumpers could make similar policies for companies bidding for federal conracts.

So, President Trumpers, how about it?  And for that matter, you too, President Obama.