From: Norm Matloff To: Norm Matloff Subject: Sloan West Coast Program on STEM Workers To: H-1B/L-1/offshoring e-newsletter A new program, the Sloan West Coast Program on Science and Engineering Workers, held its first conference yesterday on my campus, UC Davis, with a focus on H-1B. The organizer was Phil Martin, a professor of agricultural and resource economics at UCD who has done extensive research on immigration for many years. It was a very interesting and enjoyable day. You can see a list of speakers at http://migration.ucdavis.edu/wcpsew/ In most cases, links are provided to the speakers' presentations. But of course the interaction of the speakers and attendees was the most interesting portion, and I'll give a summary here of what I regard as the high points. The morning began with a welcome address by Barry M. Klein, Vice Chancellor for Research at UCD. Not surprisingly given his university administrator status, Klein gave a spirited "It's a flat world after all" speech. He even claimed that the venture capitalists prefer to fund immigrant entrepreneurs, as the latter are thriftier with their money. :-) That was followed by three presentations on employer violations of H-1B law. As many readers of this e-newsletter know, this area is less interesting to me, as actual violation of the law is quite rare, with the real problem being massive employer usage of major loopholes in the law. But the talks were good, including one by the author of the oft-cited 2006 GAO report, which found (to no one's surprise) that violation of the law is indeed rare. (See my commentary on that report at http://heather.cs.ucdavis.edu/Archive/GAOCompliance.txt) It was my turn next. Other than reviewing some earlier studies, my main focus was on the industry lobbyists' claim that most H-1Bs are "the best and the brightest" engineers and programmers needed to keep America innovative. First I extended John Miano's analysis in his 2007 CIS paper. He had shown that among the LCAs (the applications submitted by employers seeking permission to hire an H-1B), the vast majority were for the lower two of the four wage/experience levels defined by DOL. My analysis was on the PERM data, which differs in various ways from LCAs. In the LCA data, each record typically corresponds to a real worker, but it may not. By contrast, in PERM, the data for the labor certification process involved in employer-sponsored green cards, each record is definitely for a particular worker. As I've explained before, statistically speaking, this is not really an issue, but since this technical point has been stressed by the industry in its critique of John's LCA analysis, it was important to show similar patterns exist in PERM, and I did show this. In addition, the PERM data show the nationality of the worker, which turns out to be of significance. Here are the points I made: 1. The vast majority of workers, 70% or so (somewhat more for some occupations, somewhat less for others) are in either Level I (defined by DOL as "routine tasks... limited if any exercise of judgment") or Level II ("modestly complex tasks requiring limited judgment"). CLEARLY, MOST ARE NOT HIRED AS INNOVATORS. 2. Contrary to the industry's claim that they import workers from Asia because "Johnny can't do math," the only countries with large percentages of Level IV workers--the real experts--are Canada and the UK. Of course, the fact that most of the workers are at the two lowest levels also means that they are cheap. A couple of the people in the audience (one was the immigration lawyer who spoke later, and the other was a researcher who generally takes a pro-industry point of view) didn't understand this. They said, "Well, many of the H-1Bs are hired from U.S. university campuses, so naturally they're young. That's normal." Of course, they had it exactly backwards--it's not that the industry wants students, who happen to be young, but rather that the industry wants the young (thus cheap), who happen to be students. Then I addressed the question of "the best and the brightest" directly, by looking at the ratio of actual wage paid to the prevailing wage. After all, if the industry's claims are true--that the H-1Bs are mostly outstanding talents and they are paid market wages--then their wages should be way above prevailing wage, right? So, I looked at the median of these ratios. Remember, a ratio of 1.00 means that the employers are not paying above prevailing wage at all, i.e. no "genius premium" whatsoever. The results were rather striking: 1. The median over the entire PERM set was 1.02. In other words, most employers are paying either prevailing wage or only a tiny bit above it. These workers ain't geniuses, folks. 2. Broken down by occupation, the corresponding values for software engineers and programmers were 1.02 and 1.01. Again, this demolishes the industry's "best and brightest" claim. The value for electrical engineers, 1.10, is a bit higher, but still not indicating that most of these people are Einsteins. 3. I broke things down by country here too, and again it turned out that Canada and the UK--not India and China--are the ones sending us the people of above-average talent (though not Einsteins there either). It's interesting that the immigration lawyer, Lisa Spiegel (see below), who of course greatly extolled the virtues of the H-1B program, actually seemed to accept my analysis debunking the "best and brightest" myth. She said several times, both when I spoke and later in her own talk, "OK, let's forget about best and brightest. But we need these people." After lunch, Jack Trumpbour of Harvard SEWP gave an interesting and entertaining talk on former H-1Bs who have returned to India or China. One interesting point he made was that it's a lot easier for the Indians to go home than the Chinese, if kids are involved. The Chinese kids just don't have the skills in written Chinese needed to survive in school back in China, so the parents are very reluctant to return. One interesting point of contention arose when Jack quoted managers of U.S. firms operating in China as saying that while the Chinese workers would take advantage of training courses offered by the firms, their American workers were rarely interested. Well, there it is again, that old "American slacker" canard (not Messenger Jack's fault, of course). Kim Berry of the Programmer's Guild vigorously disputed this, saying that he had always grabbed the chance to get training, and everyone he knew did the same. After some discussion, it turned out that those U.S. managers in China and Kim were talking about different things, with Kim saying he always took advantage of technical training, say in the C# programming language, while the managers Jack quoted were talking about courses in project management skills such as Six Sigma. (I myself think Six Sigma and the like have little value; see my posting at http://heather.cs.ucdavis.edu/Archive/CMMHype.txt On the other hand, I don't think formal course training in programming languages is needed either; a good programmer can learn on his own, as long as he has someone to consult regarding the "gotchas" when they arise occasionally.) Then Lindsay Lowell gave a two-part talk, first on his recent study for the Urban Institute (see my postings on that excellent work, at http://heather.cs.ucdavis.edu/Archive/UrbanInst.txt and http://heather.cs.ucdavis.edu/Archive/UrbanInst2.txt), and then on a regression analysis he did on foreign worker data. I'm not sure what to make of that latter one, as it found only a weak relationship (R-squared of 0.09) and likely has a hyperaggregation issue. Hopefully he will come out with a full paper on this work and I'll be able to comment more. Lisa Spiegel, the immigration lawyer, gave a presentation consisting almost entirely of Stuart Anderson's talking points. As I've mentioned about Anderson before: Anderson has been Mr. H-1B for years. In Fall 1997 he wrote the pivotal report for the ITAA which that lobbying group used to convince Congress to enact the first major H-1B increase in 1998. In 2000 he then became a staffer for then-Senator Spencer Abraham, and in that position Anderson authored the second major H-1B expansion enacted by Congress. When Abraham's bid for re-election failed that year (in part because of Abraham's support of H-1B), Anderson then went to a top position in the Immigration and Naturalization Service. Anderson is constantly cited by the American Immigration Lawyers Association and its foundation, AILF, and presumably gets his funding from them and/or from the industry, as he did with the ITAA lobbying group. I've shown before why Anderson's various claims are incorrect or misleading, and thus won't go into it here. Lisa also trumpeted the fact that IEEE-USA wrote its infamous joint letter with the SIA industry lobbying group. She said, "See, even this labor group supports more green cards." I asked Lisa about her point that 1/4 of high-tech startups are founded by immigrants. Since there are so many immigrants in the field, it should be no surprise that there are a lot of immigrant entrepreneurs, but the per-capita rates of entrepreneurships are almost identical for immigrants and natives. (The native rate is slightly higher.) I noted that 30% of America's small motels are owned by Indian immigrants, but that that certainly doesn't mean that without Indians we wouldn't have motels. Her reply was, "Well, where did YOUR family come from?" Jeff Wheeler, the guy from Intel, had an odd speaking style, and seemed to be evading questions, quite a disappointment though not such a surprise. Kim Berry of the Programmer's Guild gave a really outstanding talk. I had seen his slides earlier, and they were fine, but his delivery greatly enhanced the content. Here was a real victim, speaking calmly yet with contained anger at the fact that all our respected institutions--both major political parties, the business community and academia--are complicit in maintaining that sham known as H-1B. His account of hiring decision meetings in which he participated, in which qualified American applicants were repeatedly rejected in favor of H-1Bs, ought to have been videotaped; his speech would have been just as effective the Cohen & Grigsby "TubeGate" videos. Stan Sorscher wrapped it up with some dramatic graphs and some excellent food for thought concerning the growing divergence between investors and the public, between the upper 5% income class and the middle class (and the lower 5%) and so on. Rep. Zoe Lofgren was a no-show. I had been skeptical from the moment I learned she was on the agenda as to whether she'd actually show up, as I'd seen her weasel out before. Once she was supposed to be on a PBS TV broadcast on H-1B. At first she enthusiastically accepted the producer's invitation, but the producer told me that when she learned the next day that I would be on the panel with her, she told him, "Then I can't participate." When the producer asked why, she replied, "Because we don't think he's credible"--an odd and of course transparent excuse to give. Yes, there was a snowstorm in the East yesterday, and that could have been the real obstacle (Lofgren was coming from Des Moines, but the entire air travel system was probably affected). But I really don't think she ever intended to show up. All in all, it was very good conference, and I look forward to the future events in the series. Norm