Date: Tue, 2 Aug 2011 00:00:28 -0700 From: Norm Matloff Subject: CHE op-ed by Salzman and Lowell To: H-1B/L-1/offshoring e-newsletter The Chronicle of Higher Education (CHE), a kind of trade paper for academics, is running an op-ed titled "A Size That Fits All for the Science-and-Technology Pipeline" in the July 31 issue. The authors are Hal Salzman of Rutgers University and B. Lindsay Lowell at Georgetown University, whose names will be recognized by many of you. Salzman and Lowell are the authors of the 2007 Urban Institute study that showed that, contrary to the quotes from industry executives that constantly bombard you in the press, the nation's colleges and universities are producing more than enough graduates in STEM (science, technology, engineering and math) fields to fill our economic needs. I reviewed the study in http://heather.cs.ucdavis.edu/Archive/UrbanInst.txt http://heather.cs.ucdavis.edu/Archive/UrbanInst2.txt Both Salzman and Lowell have done extensive research on H-1B and related issues. Of the two, Lowell has probably specialized more in H-1B, but Salzman's work has been seminal too. The much-cited research for the 2001 NRC report, which found that employers admitted to paying H-1Bs less than comparable Americans in "lower level" jobs (the term meant ordinary jobs, not the jobs taken by MIT grads and the like), was actually conducted by Salzman. Salzman was one of the presenters at the recent DC conference I reported on here (and Lowell was one of the organizers); you can view Salzman's slides at http://migration.ucdavis.edu/wcpsew/more.php?id=6_0_7 I will not enclose the full op-ed text here, due to CHE's restrictions on sharing, but (at least for now) you can read it in full at http://chronicle.com/article/A-Real-Fix-for-Science-and/128421/ I will quote excerpts and make comments. ...international testing that shows students in Shanghai at the top of the world...[an influential report] finds the deterioration of America's competitiveness so severe that it is likened to a Category 5 hurricane...It reinforces a common worry over American students' lackluster international standing compared with those in several Asian nations and in a handful of small European nations. We believe that those concerns are overstating and misidentifying America's challenges in science and engineering, and that they are missing the real opportunities for improving the nation's education and work force. I commented rather extensively on this issue in my review of President Obama's State of the Union Address this year; see http://heather.cs.ucdavis.edu/Archive/ObamaStateOfUnion.txt I pointed out first that though I admire Shanghai for its dynamism-- constructing one or two full new subway lines every year!--those test scores mean almost nothing. First, Shanghai is the richest city in China, and thus is likely to do well, just like the upper class does well in the U.S. I cited the Atlantic Monthly interactive charts that show that Caucasian students in DC vastly outscore the kids in Hong Kong and Taiwan, two of the countries that attain among the highest math scores in the world. Second, and much more importantly, I stated that the worst thing we could do is emulate China with its innovation-quashing rote memory education system, known in Chinese as 填鴨, "stuff the duck." The Chinese government itself is trying to move away from this approach, which it knows produces uncreative graduates and, as Chinese engineering professor Chen Lixin has put it, "results in the phenomenon of high scores and low ability." I could add a number of similar quotes here, including one from none other than China's current premier, Wen Jiabao, but suffice it to say that the Shanghai test scores aren't very relevant and are clouding the real issues. Our recent analysis of Department of Education data for three decades followed students from high school to the job market. We found little in the way of overall change in students' pursuit of science-and- engineering studies or their entry into those careers over the past 30 years. We found that while a steady proportion of college students graduated in science and engineering, no more than half of them landed jobs in a formally defined core science or engineering occupation. So, given a steady supply, why do companies report difficulty in finding ideal workers? Listen carefully and it sounds as if the employers would like entry-level workers to have skills not typical of newly graduated students. Leading engineering companies seek technologists with a depth of skill in a technical area combined with a broad education across technical fields, business, and the social sciences. Colleges find it difficult to develop all of that in only four years. Here the authors are giving far too much credence to the PR the industry lobbyists have put out to support their push for expanded H-1B and green card programs. In computer science, most Bachelor's degree-level workers use at most 20% of what they learned in school. Moreover, that 20% comes largely from coursework taken in their first two years. The other courses can in principle form a valuable "culture" from which to draw, but most students (whether domestic or foreign) don't really do so. In any event, a CS degree does not consist of four solid years packed with one special technology after another. Nor does the industry want it that way. In hiring new graduates, the industry does NOT want or expect an extensive technical skill set. What it does want is really smart, enthusiastic, reliable, and above all self-starter-type, people. (Note how many of these are "soft skills," a point the authors also make.) A firm foundation is important, but not the really esoteric skill sets. At the graduate level, employers often do prefer students who have done research in some very specific field, but that is quite consistent with the nature of grad school, and does not contradict my above comments. Employers claim to prefer new grads because the older workers lack the latest skill sets. Again, that is just a pretext. As I'm fond of saying, where do those young hotshots get their up-to-date skill sets, say the Python language, cited recently by the industry as a hot skill that older engineers don't have (http://heather.cs.ucdavis.edu/Archive/Python.txt) or use of GPUs for parallel processing? Answer: The young hotshots learn those from ME, an old guy. This is not quantum mechanics or brain surgery, folks. A competent software engineer (would you want to hire an incompetent one?) can become productive in a new programming technology very quickly, through self study. As I say in virtually every posting to this e-newsletter, the core of the H-1B issue is age. Engineers over age 35 are considered too expensive by employers, both in terms of wages and benefits. While it's true that H-1B also saves employers in labor expense by enabling them to hire young H-1Bs more cheaply than young Americans, they reap even bigger savings by hiring young H-1Bs instead of older Americans. I cannot overemphasize this point. I've written extensively about it, but for a couple of quick sound bites, let me cite two concerning Intel: 1. Intel's Craig Barrett, the firm's former CEO who is in the vanguard of those saying the U.S. doesn't produce enough STEM grads, once said "the half life of an engineer is only a few years," graphically illustrating the throwaway nature of labor in the industry. 2. According to the book Inside Intel, a management consultant hired by Intel to reduce costs recommended getting rid of older engineers. So what H-1B is really about is age. Unfortunately, this point is very poorly understood even by the experts. In this week's Senate hearing, various remarks critical of the industry regarding H-1B were made (including by two senators), but the relation of the age issue to H-1B was never brought up, sadly. This point will relate to the next excerpt from the Salzman/Lowell piece: Finally, some industry lobbying groups and high-tech companies seek to augment the supply of domestic workers by importing foreign labor on temporary visas. But this confuses the purpose of those programs with the country's immigration policy for citizens-in-waiting. Immigration policy is driven by a long-term vision and a wide range of social and political objectives. The original intent of temporary-visa programs, on the other hand, was to meet short-term, not structural, labor shortages. Ensuring that labor markets are not distorted by short-term visas, which in their current form lead to a number of labor-market and social problems, is not anti-immigrant, and does not undermine the strength of U.S. science and engineering. In fact, raising the numbers of temporary visas for foreign workers during cyclical talent shortages can distort labor markets and discourage domestic students from careers in engineering and the sciences. As I sometimes mention, readers of this e-newsletter consist of programmers and engineers, academics, policy makers, journalists and so on. Yet I would guess that only a small percentage of readers in any of those categories understand what is being alluded to above. The phrase "long-term" is apparently an allusion to employer-sponsored green cards, to be distinguished from H-1B. One of the common themes among policy makers recently has been that green cards are good and work visas are bad, with the next step in the argument being to legislate some kind of fast-track green cards for international students who obtain graduate STEM degrees at U.S. universities. I'm not saying that the two authors here necessarily support such legislation, but support for it is so broad that for example at this week's Senate hearing Sen. Schumer made a statement along the lines that people on all sides of the H-1B issue support fast-track green cards. That's false. As many of you know, I strongly disagree with that kind of thinking. Aside from the point that it would flood the market with workers that Salzman and Lowell have shown are not in short supply, the salient issue is--once again--AGE. Beneficiaries of these autogreen cards would in almost all cases be young, and thus displace even more of the Americans over age 35 than are displaced now. I cannot think of a more wrongheaded way to go, and would submit that it would be even more harmful to U.S. citizen and permanent resident workers than would an increase in the yearly H-1B cap. As the authors themselves say so succinctly, The classic tried and true formulation is that supply follows demand or, less sanguinely, that depressed wages and discouraged workers result if supply outstrips demand. That comment is directly relevant to the autogreen card proposals. I will be commenting here on the Senate hearing, and on proposed legislation, sometime in the next few days. Norm