Date: Tue, 18 Jan 2011 00:39:39 -0800 From: Norm Matloff To: Norm Matloff Subject: the new GAO report To: H-1B/L-1/offshoring e-newsletter I've received several e-mail messages about the new GAO report, "H-1B VISA PROGRAM Reforms Are Needed to Minimize the Risks and Costs of Current Program," available at http://www.gao.gov/new.items/d1126.pdf Most have said, "Typical bland GAO report," referring to the understated tone typically used by the GAO, a congressional research agency. But from my point of view, the situation is far worse. Though there is material of value in the report, I am profoundly disappointed. If appendices and background material such as the legal history of the visa are excluded, over half of the report concerns the H-1B wage issue. (The remainder is on employer complaints about the program.) In other words, the GAO definitely "gets it" in terms of wages being the central issue in the eyes of the critics. Given that, why did the GAO completely ignore the key point on the wage issue--the loopholes in the statutes and regulations that enable employers to underpay H-1Bs in full compliance with the law? And they did know about the loophole issue: I was interviewed by the GAO team in September 2009. As I recall, the interview lasted about two hours, with the loopholes being the most important part of the discussion. Indeed, I recall being pleased that the head of the team had done her homework, and pointed out, even before I did, that in my writings I've emphasized that underpayment of H-1Bs is a loophole issue, not an enforcement issue. In a followup e-mail message to the head of the team after the meeting, I thanked her for understanding this absolutely key point. H-1B employers are required to pay the "prevailing wage," a legal term. Due to the loopholes, the legal prevailing wage is typically well below the true market wage. I explained to the research team exactly how the loopholes work, such as the fact that the OES wage does not take into account "hot skills," which on the open market can command wage premiums of 20% or more; in other words, just that one loophole is already enabling a substantial underpayment, and there other various others. And this point about skills is accentuated by the fact that the GAO report repeatedly quotes employers as saying they hire H-1Bs because the latter have skill sets that Americans don't have. The GAO team didn't have to take my word for this. They could have read the statutes and regs themselves, and then consulted with immigration lawyers for details. Perhaps most importantly, they could have heeded the GAO's own 2003 report (which I'm sure I pointed out to them), which stated Some employers said that they hired H-1B workers in part because these workers would often accept lower salaries than similarly qualified U.S. workers; however, these employers said they never paid H-1B workers less than the required wage. That passage illustrates that the legally required wage (the prevailing wage) is below the true market wage. The research team also was probably aware of "Tubegate," in which a prominent law firm discussed loopholes in prevailing wage. (They were discussing green cards rather than H-1B, but it's the same loopholes.) I also pointed them to former industry CEO Vivek Wadha's writings, in which he states that loopholes are used to underpay H-1Bs. Once again: The central point of criticism of H-1B is wages, and the key point on wages is the legal loopholes. For the GAO to ignore this was unconscionable. It should be noted that it is especially unconscionable in that the GAO claims that it "examined...how well H-1B program requirements ensure that U.S. workers are not displaced or otherwise disadvantaged by the program." And on the contrary, they devote many pages to the enforcement issue, which again is a NON-issue; in fact, the GAO's 2006 audit found the compliance rate to be quite high. Well, of course it's high--why should an employer risk debarment from the program when he can already make out like a bandit just by using the legal loopholes, perfectly safely? To add insult to injury, the report then proposes that compliant employers be "rewarded" by subjecting them to a more lax procedure, a surreal "reward" in view of the above. By the way, when I spoke at a recent conference on immigration at UC Berkeley, I made this same point about prevailing wage, as I always do. During the Q&A, there was a question from Angelo Paparelli, a very prominent immigration lawyer and pro-H-1B political activist. He said, "I agree with you that the prevailing wage is below the market wage, but that problem is solved by the requirement that the employer pay either the prevailing wage or the actual wage, whichever is higher." The "actual wage" he was referring to is the average wage of workers in the same job with the same employer. I replied that there are loopholes there too. For example, the H-1B's job can be defined in such a way that the the H-1B is the only one with that job, so his wage is the actual wage; indeed, the DOL Web page notes this. Or, the employer can simply claim that the other workers in the same job are making more because they are better qualified, an assessment that DOL would find it tough to dispute, given that the employer is the one holding the cards, knowing technology. Or, if all or most of the workers in that job are H-1Bs, the whole idea of actual wage becomes moot. Furthermore, Paparelli's distinction between prevailing wage and actual wage is rather pointless, in that most employers treat the two as the same. In the LCA data, for instance, the employer's prevailing wage and actual wage are typically the same. In fact, Paparelli himself said in his blog, at http://www.ilw.com/articles/2009,0413-paparelli.shtm the following: The dozens of pages of H-1B regulations published by the Labor Department require that H-1B workers be paid the "required" wage - a term defined as the HIGHER [emphasis is Paparelli's] of the prevailing wage in the geographic area where the work is performed or the actual wage paid to similarly qualified workers in the same job at the employer's worksite. In most cases, employers merely use the prevailing wage for the particular job that the Labor Department provides. So Paparelli's prevailing wage/actual wage distinction is a non-issue. The team also did a simple analysis comparing H-1B wages with those of U.S. workers. I'm sorry to be so critical, but the fact is that here again they dropped the ball, because their analysis does not take education and other variables into account. This failure is remarkable, because they did emphasize the fact that H-1Bs as a group have more master's degrees and so on than the U.S. workers (pp.89-90). So, when the team found that in the younger age groups, the H-1Bs are paid about the same as the Americans, THAT MEANS THE YOUNGER H-1Bs ARE UNDERPAID RELATIVE TO YOUNGER AMERICANS WITH THE SAME EDUCATION. Granted, they themselves recognized that the data they used for H-1B wages did not include the education variable or other big factors, but there are other data sets they could have used that give both visa status and education (Nat. Survey of College Graduates), and the CPS data they used does enable proxies for H-1B status, as I've shown in my own analyses. Well, then, what did the GAO do well? They covered the age issue fairly well, and in more depth than I would have thought. There is a considerable discussion on this (though of course not going further than simply reporting "Many labor advocates believe..."), and Figures 16, p.89, is wonderful. They show age distribution histograms for H-1Bs in dark blue, and then display the histograms in upside-down mode for the Americans below in light blue. Well, guess what--in all but one case, the two histograms are almost like two pieces of a jigsaw puzzle that fit neatly together. In other words, if you move the light blue histogram up, just above the dark blue one, you more or less get a rectangle! A very stark illustration of the fact that the H-1Bs are young and the U.S. workers old, an excellent pictorial for those who contend that above all, the H-1B work visa is really about age--it gives employers a way to avoid hiring older U.S. workers. Indeed, you can see in the figures the cutoff point at age 35, the number I always use in my writings. The report also discusses at some length the point first brought up in John Miano's research, which showed that the vast majority of H-1Bs are hired into very low-level positions. In fact, most are at Level I, officially defined by the DOL as needing a "basic understanding of duties and perform routine tasks requiring limited judgment." This belies the industry lobbyists' claims that H-1Bs are hired because they are experts that can't be found among the U.S. workforce. The GAO stops short of concluding this, though, using hedging language. Though I may be biased, I felt that the GAO did not use hedging language when reporting the concerns of the mainstream industry, especially startups. The GAO accepted without any question whatsoever the claims that startups need someone to be productive the day they're hired (impossible, for many reasons), that Americans can't learn new skills quickly on the job, etc. The only segment of the industry that the GAO seemed to be critical of was the bodyshops, referred to by the GAO as "staffing agencies." As I've mentioned many times in the last two years or so, in certain circles of government and academia, a "conventional wisdom" has developed in which Tata Consultancy Services et al are considered the Bad Guys, while Intel etc. are viewed as the Good Guys. This image has actually been carefully fostered by Intel etc., and unfortunately has been fed by the Programmers Guild and other labor advocates. As I've explained before, this image is false; the mainstream firms abuse the H-1B program just like the bodyshops do, albeit hiring a higher class of worker. And of course, concomitant with this image is that the bodyshops are violating the rules, which in fact they generally are not. Unfortunately, the GAO appears to have bought into this view that the staffing agencies are the bogeymen. I was pleased to see that the GAO did mention the awful situation regarding science post docs, and did imply that H-1B plays a role. On the other hand, they failed to note the NSF's advocacy of H-1B as a way to flood U.S. university grad programs with foreign students in order to keep PhD wages down. The GAO recommended that there be a national Web page showing which employers intend to hire H-1Bs in which jobs, so that American workers can apply. I've suggested this too, but it would be of limited usefulness, since employers would overdefine job requirements and because they can legally reject qualified Americans anyway. On that latter topic, the DOJ's letter responding to the GAO, included in the report, recommends legislating a labor market test requirement for H-1B. Sadly, the DOL did not agree, presumably again because they believe the startups' claim to need speedy response to their H-1B applications. Overall, though the report does contain some positive material, they fail on the most important point, concerning the loopholes, which form the lynchpin of H-1B. This is inexcusable. Norm