Yet Another Misleading Report

There is no shortage of pro-H-1B research reports sponsored by pro-H-1B organizations. The latest is The Demographics of Innovation in the United States. As with the others, this one has the usual characteristics:

  • Reliance on gross figures, not broken down or compared to the relevant numbers.
  • Absurd comparisons of educated immigrants to the general U.S. population.
  • Numerous citations of pro-H-1B research papers.
  • Selective quotation of other papers.
  • NO references to papers with contrary findings.

Those last three bullets are especially egregious in this ITIF report, as it claims to present a comprehensive “[research] literature review.”

I will not go into a detailed critique of the report, and will just focus on what the report hopes will be its “headline grabbing” finding:

The study finds that immigrants comprise a large and vital component of U.S. innovation, with 35.5 percent of U.S. innovators born outside the United States.

Is 35.5% a lot? Not when you take into account the relevant bases of comparison, such as:

  • More than 60% of Silicon Valley engineers are immigrants.
  • Over 50% of computer science PhDs granted in U.S. universities are earned by foreign students.
  • So that 35.5% figure would suggest that the immigrants are underperforming relative to the natives, not outperforming them.
  • On the contrary, the per-capita figures show that immigrant computer scientists and others in STEM are LESS likely than U.S. natives to have filed a patent application or even be in R&D.

So that 35.5% is a perfect example of “how to lie with statistics.” In many cases the “lies” are simply mistakes arising from ignorance, but whether deliberate or not, the harm done is the same.

I’d say that it’s likely that some senator will cite this ITIF report in tomorrow’s Senate hearing.

6 thoughts on “Yet Another Misleading Report

  1. Norm did you not bother to read the report and in particular figure 18 and table 16 which clearly show that immigrants are important in US innovation? If you are going to accuse us of lies at least have the courtesy to read and report what we say accurately, which i assume is what academic reaearchers are committed to do


    • Thanks for your comments, Rob, but I don’t think those two items back you up at all. Table 16 shows NO overrepresentation of immigrants. On the contrary, you and your coauthors yourselves say,

      The percentage of non-naturalized immigrants producing innovations in our sample matches the percentage of non-naturalized immigrant scientists and engineers. However, non-naturalized European and Oceanic scientists and engineers are more likely to have produced innovations, representing 6.9 percent of responses, but making up just 3 percent of U.S. scientists and engineers.

      So, the first part of those remarks of yours shows, as I said, that the immigrants are not more prone to becoming innovators. Given the demonstrable displacement (indirect and direct), which even your reference Giovanni Peri concedes (did you know that?), that means NO GAIN FROM IMMIGRATION.

      And in fact there is a LOSS. It’s fine to look at the creme of the crop, and I myself look at awards in my EPI paper, where I find, “…foreign students earned the dissertation award roughly in proportion to their presence in the doctoral student population, again indicating that the foreign students are similar to, rather than more talented than, their American classmates.” And I have always strongly supported facilitating the immigration of “the best and the brightest.” But clearly it is wrong and misleading to look only at the elite.

      If you look at the immigrant STEM population as a whole, things change markedly, with a definite trend of the immigrants being of somewhat lower quality compared to natives. The immigrants have a lower rate of patenting, are less likely to work in R&D (read “innovation”), earn degrees at less-selective universities and so on. Again, given the indirect and direct displacement of Americans, IMMIGRATION IS CAUSING A NET LOSS IN INNOVATION.

      I didn’t accuse you of lying. I gave you and others an out by saying your errors may be due to ignorance, meaning lack of information on methodology, structure of the industry and so on. It’s important, for instance, to adjust for job market sector, such as the fact that Americans tend to go more into government positions, in which they typically don’t do research.

      The second part of your quoted comment above, that the immigrants from Europe and Oceania (presumably referring to Australia/New Zealand) is important in a way that you failed to mention: The immigrants from these countries tend to come — i.e. are filtered for immigration — after they have established themselves as outstanding, say 10 years after their degree, as opposed to the vast majority of H-1Bs being hired right out of school. Indeed, you make the point that these innovators are older. So, if your message is to expand the H-1B program (I assume you won’t deny this), then your data are inconsistent with that message. And by the way, David Hart’s entrepreneurship study shows the same thing.

      I did accuse you of assembling a blatantly one-sided bibliography, citing NO research contrary to your theme. This is absolutely inexcusable.


  2. Norm, it’s amazing what you can find on the Internet these days.

    Apparently, Maria Klawe, who you know is the President of Harvey Mudd College, and former board member of Broadcom, who approved its sale to Avago, leading right away to the offshoring of at least 700 highly paid Masters and PhD level STEM positions, and who also received the 2014 Anita Borg “Women of Vision ABIE Award for Leadership”, for her supposed “fearless pursuit of closing the gender gap in technical education”, also testified in front of none other than H-1B visa evangelist, Senator Maria Cantwell of Washington State, in 2013.

    Here’s the link to Maria Klawe’s testimony in front of Maria Cantwell:


    “To me the most important language today is computer programing language,” said Cantwell at today’s hearing. “Should we look at incentives at the federal level to encourage states to make something like C++ or Java as part of a one year curriculum requirement for high schools or incent high schools to do that? So more and more people are exposed — just as I was forced to take typing — get people exposed to what really is going to be the language of the 21st century.”


    Dr. Klawe replied: “I think there are many initiatives but the one thing that’s not there in most places is the requirement to take some computer science either in middle school or in high school. And we need it. So yes, that would be a wonderful thing to have happen at the state level and any help from the federal government would be very welcome.”,

    No insider dealing there, right? Wasn’t Maria Klawe on the Microsoft Board at the time? Microsoft, based in Redmond, State of Washington, where Cantwell is a senator.

    And now that the entire country is bending over backward to teach C++, Java, and HTML in middle school, while often sacrificing their math, science and language programs. Isn’t it so great that Maria Klawe (girls need special computer games, and could never otherwise be interested in science and engineering, don’t you know?) and typist Maria Cantwell, have now re-vectored the middle school curriculum of the entire country?

    Maria (Cantwell) I’m so glad you learned to type in high school. I also heard through the grape vine that you learned other skills in high school (not directly related to your job) which, in addition to typing, helped you “move up.”

    Of course, the problem with those displaced American engineers and computer scientists was that they didn’t learn to program C++, Java and HTML in grade six. I’m sure that’s why they deserved to be booted out of the workforce at age 35.

    Anything to create the illusion of a continued need for the fraudulent H-1B program, right? Even if it means distorting the middle school curriculum of the entire country and wasting precious education dollars.

    Oh, also, Maria Klawe, having been on the board of Microsoft since 2009, could not possibly have known about the gender pay gap there. Not for five years until her sudden epiphany in 2014.

    Apparently, Maria Klawe, is attending a “Future of Higher Education” symposium at UC Irvine this Friday, February 26th:

    Ralph Cicerone, President of the National Academy of Sciences, will also be there.

    You can bet your bottom dollar that the H-1B visa, and its devastating impact on STEM career, will not be on the agenda. (Oh, and neither will that other hot button item (#astrosh)).


    • Klawe proposed (and implemented, I believe) teaching introductory programming in Python rather than C/C++/Python because “women would find it easier.” This is appallingly offensive and false (though I do agree that EVERYONE should start with Python).


  3. > I will not go into a detailed critique of the report, and will just focus on what the report hopes will be its “headline grabbing” finding:
    > > The study finds that immigrants comprise a large and vital component of U.S. innovation, with 35.5 percent of U.S. innovators born outside the United States.

    The report succeeded in pushing its “headline grabbing” finding. I became aware of this study from a CNET article at titled “Well-educated immigrants are biggest innovators in US tech”. The article states:

    > It found that 35.5 percent of them were immigrants. That far exceeds the proportion of first-generation immigrants in the US population, which stands at about 13.5 percent, according to the report published Wednesday by the Information Technology and Innovation Foundation.

    Oh my gosh, 35.5 is almost three times 13.5! Let’s bring in as many of those geniuses as possible!

    Table 16 and Figure 18 mentioned by Rob Atkinson in his comment show that the percentage of foreign-born innovators are actually just in line with their percentage among scientists and engineers (table 16) and Ph.D. holders (figure 18). However, the first finding given at the start of the second paragraph of the Abstract is:

    > The study finds that immigrants comprise a large and vital component of U.S. innovation, with 35.5 percent of U.S. innovators born outside the United States.

    The 13.5 figure for immigrant representation in the entire population is only mentioned in Table 8. Page 29 gives it as 13 percent of the U.S population. Yet this wrong-headed comparison of 35.5 and 13.5 is what the media latched onto. Part of the reason for this may have been that ITIF introduces this “groundbreaking” study on its web site at with the statement:

    > Highly educated immigrants play an outsized role in driving technological progress in the United States, while women and minorities are significantly underrepresented among the country’s innovators, the Information Technology and Innovation Foundation (ITIF) reported today in a first-of-its-kind study.

    As it’s first finding, it states:

    > More than one-third of U.S. innovators (35.5 percent) were born outside the country, even though first-generation immigrants comprise just 13.5 percent of the total population. Another 10 percent of U.S. innovators had at least one immigrant parent.

    This seems to be the pattern with most of these studies. The study itself lists some of the limitations and uses somewhat more careful language. But then the Summary and Conclusion, highlighted excerpts, and introductions of the study become more and more strident. The media then just takes it several steps further.

    Speaking of the conclusion, it contains the following assertion:

    > One major factor holding the United States back in this regard is the lack of an adequate workforce in STEM fields.

    I have not read the entire paper but I did not see that they give any evidence for the so-called STEM shortage. In fact, as can be seen by the articles at , this is, at the very least, a hotly contested assertion.


  4. The paper’s reliance on patent grants is naive, and must undermine its findings.

    1. Large corporations file patents routinely as a business strategy, based on whatever the labs have been working on for the past six months. It’s a standard part of the duties of the legal department. Those patents do not necessarily represent anything innovative.

    2. Because of the above, it’s common for junior staff at large corporations to have their name attached to patents. That doesn’t mean the staff are particularly innovative or even very good.

    3. The paper argues that its reliance on 3-nation patents filters out weak patents. That’s wrong. For large corporations, overseas filing is handled by the overseas offices. It doesn’t cost three times as much, as the paper claims. The main work is the initial drafting.

    4. Reliance on patents excludes a lot of important software innovation.This is particularly the case for 3-nation patents, due to Europe’s opposition to software patents.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s