Concerning H-1Bs, the report omits the following:

• The finding in the DOL audit that 19% of the H-1Bs were not even being paid what their employers had stated on the visa application.
• The finding in the INS audit that 21% of the H-1B applications were found to be fraudulent, and 29% more "were either probably or possibly fraudulent."
• The finding in the DOL audit that 33% of the H-1Bs left their jobs within one year of getting a greencard, illustrating the indentured servitude nature of the H-1Bs.
• The finding by the GAO that the H-1B program fails to protect either American workers and the H-1Bs themselves.
• The DOC finding in 2000 that H-1Bs now account for 28% of all IT requiring at least a Bachelor's degree.
• The finding in the study by Papademetriou and Yale-Loehr (Carnegie Institute for International Peace and Cornell University) that employers who were sponsoring foreign nationals for greencards were paying them salaries which on average were more than 20% below normal, in violation of the maximum 5% gap mandated by law.
• The finding by UCLA professor Paul Ong that foreign-born engineers made on average 33% less than comparable natives nationwide.
• The finding by UCLA professor Paul Ong that foreign-born engineers made on average 33% less than comparable natives nationwide.
• My finding that recent-immigrant programmers and engineers in Silicon Valley were on average being paid 15-20% less than comparable natives.
• My data showing that employer usage of H-1Bs has grown much faster than the growth of jobs in the field--including during the early 1990s, when there was a recession and massive defense downsizing. Concerning older programmers and engineers, the report omits the following:
•  The finding by American University professor Laura Langbein that the time needed to find another job for laid-off engineers increases three weeks for each additional year of age.
• The Informationwek survey showing that only 2% of hiring managers would seek a worker having more than 10 years of experience.
• The Network World survey data showing that the younger the manager, the less likely they are to hire people over 40.
• My finding that 20 years after graduation, only 19% of CS graduates are programmers, while 57% of civil engineers are still in the field at that time. (And my related data showing that it is NOT due to programmers becoming managers; the civil engineers actually have a higher rate of going into management than do the programmers.)

• The 1999 finding by the DOC that there is not enough data to determine whether a labor shortage exists, reversing its 1997 finding of a shortage. (The NRC report mentions the 1997 finding but not the 1999 one!)
• My data on hiring rates and interview-to-hire rates.
• The results of John Templeton's "sting" operation, in which he sent resume's of African-American programmers and engineers to Silicon Valley firms which hire H-1Bs, and got zero response.
• My data on the (small) number of universities at which major employers recruit new CS graduates.
• The finding by Clifford Adelman of the Dept. of Education that large numbers of non-computer science majors take at least mid-level courses in CS, thus providing a large pool of potential programmers.
• The EPI data showing that CS enrollment has risen and fallen in almost perfect correlation to conditions of the labor market, thus contradicting the committee's claim that American youth have neither the background nor the interest to study CS.

• The committee noted my point that the industry, in spite of claiming to be "desperate" to hire, actually has very low interviewing rates, very low hiring rates, and very offer/interview ratios (the last being the percentage of interviewees who are made offers). But they immediately dismissed all that by citing arguments made by industry, e.g. chiefly that one applicant many apply to many companies but can't be hired by all of them.

  The committee asked me about this during my presentation to them in October 1999, and I rebutted the industry argument by noting that even the offer/interview ratios are low. In other words, the key point is that the employers are VERY PICKY as to whom they hire. And in turn, the reason employers say they are picky is skill sets; if the employer has a project using the Java programming language, the employer will automatically reject all applicants who lack work experience in Java.

  I emphasized this not only during my presentation to the committee but also in numerous subsequent e-mail messages, as well as during phone conversations with Herb Lin. Yet my counterargument was not mentioned by NRC in its report; they said (I'm paraphrasing) "The low hiring rates are due to applicants sending multiple resume's, and that's that!"

  This is in spite of the fact that I urged the committee to check things for themselves, simply by calling a few employers to verify my points that (a) the employers reject the vast majority of their programming applicants WITHOUT EVEN AN INTERVIEW, (b) the stated reason for rejection is that most applicants don't have work experience in a specific skill, say Java, and (c) they are not interested in programmers who have taken a class in Java or learned it on their own, insisting on actual work experience. Herb Lin promised to try this, but as far as I know he never did.

  Of all the flaws in the report, this one is arguably the worst. The reason is that the low rates and ratios show that SKILLS COMPRISE THE CENTRAL ISSUE IN THE ALLEGED "LABOR SHORTAGE." One can argue whether the employers' insistence on specific skills is sincere or warranted, but the central role of skills is FACT, as the employers themselves will readily admit if you only take the trouble to ask them.

  In other words, if there is a "shortage" (it now becomes a matter of semantics), it is a shortage of skills, not a shortage of people with programming experience. This has implications for almost every issue addressed by the committee (e.g. if the educational system were to produce more programmers, all that would accomplish is give employers more people to reject each time a new programming language comes into use). So again the role of skills is absolutely central. Yet even though the report does mention this in passing (including an amusing though highly unsurprising anecdote by Herb Lin about how the response rate to his resume' changed dramatically by adding words saying that he does NOT know Java), the report as a whole ignores the point. The pickiness of the employers is an undisputed FACT, key to the entire analysis, and thus the low hiring rate and offer/interview ratio data should have not only been presented but also HIGHLIGHTED, as one of the report's displayed Boxes.

  * I had stated that any competent programmer can become productive in a new programming language within a couple of weeks, and thus the employer insistence on skills is unwarranted. I even gave quotes from industry people, including Bill Gates, making this point. Yet the report insists that, for example, a programmer in the C language (the standard of the past 15 years) could NOT do well in Java so quickly, because of Java's "object oriented programming" (OOP) nature. They say that OOP is more than just a change in language syntax; it represents, they say, an "abrupt change" in the mental processes involved, and thus it takes a long time to become productive in Java. Yet they offered no empirical evidence for this claim, and in fact it is a bunch of baloney.

  I've been programming since I was 17 years old, and have had to keep up with numerous "abrupt changes" in programming methodology, and all I can say is, Programming is Programming is Programming. It's all the same thought processes. Again, my quotes of Gates et al back this up, and in my "Debunking the Myth of a Desperate Software Labor Shortage" paper, I explain that most programmers work in teams, and when a Java novice joins a Java project, the data structures are already laid out, so that the novice is forced to program in OOP form, and thus gets used to it quickly and naturally. But none of this was mentioned in the NRC report, because it would destroy the underpinnings of industry's claim to "need" H-1B Java programmers. On the contrary, the report presents an exaggerated and incomplete view of the Boehm study on this issue (though not necessarily applied to OOP), spun to justify the industry insistence on Java or whatever other skill.

• Though the committee concedes that there is some exploitation of the H-1Bs, and does indeed call for measures to ameliorate the H-1Bs' indentured servitude, the committee denies that the wage exploitation is pervasive, especially among large employers. 

  Yet the laws of economics show right away that the indentured servitude implies wage exploitation (and thus that the pervasive indentured servitude, admitted by the committee, implies pervasive wage exploitation): If one cannot leave one's job, then one loses the chance to find a higher-paying job in another firm, or to negotiate a higher wage with one's present employer by threatening to leave. In addition, the committee ignored the many instances in H-1B Hall of Shame's H-1B database of large firms, e.g. Qualcomm, hiring H-1Bs at below-market wages.

• The committee parrots the industry line that exploitation "must" be rare, since the DOL has found rather few cases of it. This is an egregious misrepresentation, since (a) the H-1Bs are in no position to complain about exploitation, due to fear of recriminaitons, and (b) the DOL itself has said repeatedly that Congress tied DOL's hands so that it has very little power to investigate violations.
• The committee says that H-1Bs comprise AT MOST 10% of the Category 1 IT workers (programmers and the like). This may be true but fails to tell the reader that among NEW hires the rate is now 28%, according to the DOC. 

  Moreover, this fails to account for the labor-market impact of all the former H-1Bs who now have greencards or are naturalized U.S. citizens.

• The committee actually gives a very good account of why it is infeasible for many employers to ship programming work abroad. Yet the committee flatly dismisses the idea that employers hire H-1Bs as cheap labor, "because" if they wanted to save money they would ship the work abroad! The fact that study director Herb Lin, a PhD in physics, would make such an illogical statement boggles the mind. The fact that employers have decided that the work must be done in the U.S. certainly does not imply that they don't want it done as cheaply as possible in the U.S.

  The next time Herb asks his boss for his raise, his boss ought to say, "You've elected to work in the nonprofit sector, where salaries are lower. Therefore you are not interested in a higher salary." Just as Herb would like to maximize his salary subject to the constraint of working in the nonprofit sector, software industry employers would like to minimize salaries, subject to the constraint that the work be done in the U.S.

• The committee restricts its study of older workers to legally-defined age discrimination. This is in direct violation of the mandate Congress gave the committee: Congress asked NRC for a GENERAL assessment of the problems faced by older workers, NOT only legally-defined age discrimination. (Even the report, in its preface, said that Congress asked them "to investigate the status of older workers in the information technology field.") Moreover, even though the committee does look somewhat at the broader issue and finds some problems (older workers are more likely to be laid off, get paid less in their new jobs, etc.), it still does so in the context of legality, describing these as legitimate business decisions which do not violate age-discrimination laws. (Moreover, it says that IT workers find new jobs as quickly as the younger ones, ignoring my  point that the jobs they find are typically in non-IT fields, i.e. they are forced out of the field.)

  Presumably this brilliant maneuver of focusing only on the legal definition of age discrimination came from committee member Ira Rubinstein, associate general counsel at Microsoft, who would have been only too happy to provide "advice."

  What is most outrageous about all of this is that the committee is basically saying that since it is legal for employers to shun older American workers, who are perceived as more expensive, in favor of the younger and thus cheaper American workers, it is also fine for American employers to shun older American workers and hire young H-1Bs instead. Even the industry lobbyists have never gone this far in their public statements.

•  Even worse, the committee said that even if employers were to hire older programmers, this would have only a tiny impact on the shortage/tightness. Their "analysis" in this regard, depended in part on an argument which I had repeatedly warned them was fatally flawed, regarding unemployment rates among older programmers. I had noted that since, 20 years after graduation, 81% of computer science graduates (and presumably graduates of other fields who became programmers) are no longer programmers, they do not show up in unemployment rates for *programmers*, and thus such rates should not have been used.

  The NRC analysis also assumed that only 25% of those who leave the field do so because they were forced to, which is contradicted by the report's own admission that the IT industry shuns the older workers because the rapid-moving technological change passes them by.

  Finally, what is especially irksome about this is the number of older people they presume to have programming training/experience. The committee guesses this to be 357,000, just from extrapolating from the number of young programmers. The fact is that the CS graduates alone should number at least 500,000 (average of 20,000 per year for 20 years), and given that only 25% of programmers have CS degrees, that 500,000 could be well over, say, 1,500,000. Yet, NRC says that out of all of these people, only 1,010 could be lured back into the programming field (assuming the employers consented), which is ridiculously low even if the base were only 357,000. Didn't anyone bother to ask, "Does this pass the sanity test?"?

• The committee says that the job market must be especially "tight" (they play semantic games with the word "shortage") in places like Silicon Valley, due to the fact that salaries are higher in Silicon Valley than the national average. This of course ignores the fact that the San Francisco Bay Area simply is an expensive place in which to live. San Francisco bus drivers make $57,000 a year; does that mean there is a "shortage" of bus drivers? The fact is that the Bay Area is a major immigrant-receiving area of the country, which drives up real estate values and thus the overall cost of living. I live 50 miles from Silicon Valley, and yet the average 4-bedroom home in my (middle-class) city now costs nearly $600,000.
• In one of the footnotes, the committee did note the existence of the DOL audit, but immediately dismissed the audit by citing criticisms of it concerning the congressional intent of the H-1B program. Not only did the committee fail to note that that criticism came from industry lobbyists, it also failed to note that the lobbyists did NOT dispute key DATA in the audit, such as the 19% figure on H-1Bs who were not getting paid the salaries promised in the visa application.
• The committee does concede that even if the employers were to give full pay parity to the H-1Bs, the presence of the H-1Bs in the labor market would suppress wage growth in this field. (The committee may cite this as "evidence" of their open-mindedness, but even the industry lobbyists have made such statements, so it isn't much of a concession.) Yet they ignore my point, made repeatedly, that even if employers were to give full pay parity to the H-1Bs, they would hire the younger H-1Bs instead of the older Americans.
• Back in 1997, the industry lobbyists adroitly started pushing the Education Button, saying that American school kids fare poorly in comparison to South Korea et al in standardized test scores in math and science (TIMMS project), and that supposedly is why university enrollment in CS has gone down. In 1998 they were exposed as having suppressed the information that CS enrollment had been skyrocketing since 1995, yet they have continued to use education as a diversionary tactic in this manner ever since then.   

   The whole thing is a red herring in the first place, since computer programmers don't use math and science, a point which the committee chose to ignore. (As the report itself points out, most high-tech H-1Bs are programmers, not engineers, but let's continue on this anyway.)

  They also ignored Prof. Richard Rothstein's data showing that just as many university students major in some subfield of the overall science and engineering area as ever. So what is the supposed problem with kids not knowing math and science?

  They ignored the data showing that the U.S. has the second-highest number of engineers per 1,000 workers in the world (the highest is Israel), and that the U.S. rate is more than double South Korea's, that putative paragon of math-savvy kids. As to the TIMMS study, there has been extensive rebuttal in the literature, none of which is cited by NRC.

• In one of the few parts of the report I really liked, the committee does an analysis showing that there are large financial disincentives for American CS students to go on to graduate study after getting their Bachelor's degrees. The salary differentials just don't justify it. They point out that it would take 50 years to recoup one's salary losses incurred by studying for a doctorate instead of working in industry with just a Bachelor's! I might quibble with some of their assumptions and numbers there, but I highly agree with the conclusion. This is important, because the industry lobbyists are always citing the high proportions of foreign students getting PhDs in CS and engineering, implying there is "something wrong with" American students.

  Yet the committee errs badly even here. First, they omit the Hudson Institute data showing just how small the salary premium is for a PhD in CS, relative to fields like economics and political science.

  More importantly, they refuse to tell the reader that the National Science Foundation actually PLANNED it this way. From NSF internal memos obtained by Eric Weinstein of Harvard University, the NSF PLANNED to bring in foreign students to the U.S. in order to keep PhD salaries down, and ADMITTED that this would discourage American students from pursuing graduate study, while foreign students would be attracted because U.S. study is a steppingstone to get a greencard.

  Note too that I have always said there is really no good technological reason to pursue a graduate degree in CS either. The committee, at various points in the report, says otherwise, yet never offers any evidence of this.

• Though the report concludes merely that there is a "tightness" in the labor market rather than a "shortage" of workers, this is just a bald semantic game. The report uses language akin to "shortage" throughout; for example, the phrase "the demand is outstripping the supply" appears several times.
• Though the report does make some pro-labor concessions at various points in the text (sometimes in footnotes), most of them do not appear in the Executive Summary. This is again a point of bias, as only the most highly dedicated policy analysts will ever read the full text.

  Again, the committee was aware of these arguments.

• At various points the committee (a) admits that employers want to hire cheaper workers (Americans and H-1Bs), (b) employers insist on hiring only programmers who have the most up-to-date language skills, (c) employers are not willing to hire older programmers and retrain them in new programming languages (per NRC, partly because they then would demand more money!), and (d) employers insist on actual work experience anyway, claiming they need people who can "hit the ground running."

  The implication of all this is that the allegedly "temporary" labor shortage/tightness--temporary until the educational system can produce more programmers--will actually be PERMANENT. Even if one assumes that the number of CS graduates is not enough now (a point I dispute), items (a)-(d) above show that it will ALWAYS be the case that most programmers don't have the most up-to-date skills in this rapidly-changing field, no matter how many programmers the schools produce. Producing more programmers would simply give the employers more people to reject.

• Remember, the NRC actually *endorses* these attitudes among the employers concerning (a)-(d), on the grounds they are "legitimate business decisions." That means programming will continue to be a short-lived career for most people in the field in the future. Once word gets around on this point among university students, how many of them will want to study CS? The committee says it wants to increase production of CS degrees, and yet its own findings imply a big disincentive to go into this field.
• As noted above, the committee found that there are huge disincentives for American CS graduates to pursue a Master's or PhD in the field. This leads to foreign students dominating those degrees, especially the PhD. The industry lobbyists have claimed that this is a reason why many employers hire H-1Bs. This is an exaggeration, since only a minority of H-1Bs even have a Master's, let alone a PhD, but nevertheless we should address the question of that minority. The committee is not facing up to the point they raised: Does it mean that they WANT graduate programs in CS and engineering to be dominated by foreign students? If not, what can be done about it? How about cutting down the number of foreign students, so that PhD salaries can rise to market rates, which the NSF worked to suppress? The committee is silent on this.

In practice, what that means is that I did everything necessary to keep the process going, from arranging logistics to drafting pieces of the report," Lin said. "I am a paid staffer, and the committee members are volunteers. In any dispute between the committee and staff, staff loses. This is not a process in which staff drives the report. The committee drives the report."