Open Letter Regarding Raising of the H-1B Cap (Part 3)

To Whom It May Concern:

I am writing this in response to recent developments regarding H-1B visas. Some of those developments are summarized in a recent blog post by Norm Matloff titled "America's Labor Party" at this link. Sadly, the title refers to the Republican party. I believe that, like me, Norm considers himself a liberal on most issues. It is therefore mystifying that the most visible support for native tech workers these days is coming from a few members of the Republican party.

I suspect that one problem might be that many Democrats blindly accept the most visible studies on the matter even though many of them are funded by lobbying groups. For example, here are two related examples involving the White House. The first comes from a report from the Executive Office of the President at this link which states the following on page 7:

"Moreover, studies indicate that every foreign-born student with an advanced degree from a U.S. university who stays to work in a STEM field is associated with, on average, 2.6 jobs for American workers."

The second is from a post by Todd Park on the White House Blog at this link which states the following:

"Every foreign-born graduate with an advanced STEM degree is associated with, on average, 2.6 jobs for American workers."

In fact, the 2.6 figure comes from a single working paper titled "Immigration and American Jobs" at this link. It was written by economist Madeline Zavodny and funded by two organizations, at least one of which funds numerous studies with the same basic conclusions. I wrote to Zavodny and obtained her data and programs so that I was able to replicate the key findings of the study. You can see the results at http://econdataus.com/amjobs.htm. That page lists the key issues with the study and links to details on those and other issues. Following are a few of them:

The native worker employment rate used by the study incorrectly counts those not in the labor force, both retired and other, and the self-employed as unemployed.
California, which has by far the most such workers, shows a negative correlation, counter to the study's findings. This is likewise true of a number of other states.
The selection of 2000-2007 is critical to the slope of 0.0045 on which the 2.6 jobs number depends. For example, 2002-2009 gives a slope of -0.0020 which would imply that such workers were associated with a LOSS of jobs for that period.
Correlation does not imply causation though the study implies this repeatedly as explained at this link.

The replication illustrates how such studies entail a large number of choices. These many choices allow the researcher to, purposely or not, make choices that help to reach certain conclusions. For example, it was only by adding the state and year as variables to the regression that a positive correlation was achieved. For this reason, it's not enough to ensure that the calculations are correct. The study really needs to be replicated by non-partisan researchers so that the effect of these various choices can be estimated. Put more simply, it's not enough to ensure that the study is "true". It needs to be analyzed enough to ensure that it is reasonably close to the "whole truth" or, at the very least, is not misleading.

I, of course, don't expect for you to take my word as to the "whole truth". I have posted all of the programs that were used for this analysis on GitHub at https://github.com/econdataus/amjobs. I would strongly advise you to pay no attention to any study whose authors will not release their data and calculations so that the study can be replicated. For example, I wrote an analysis of a Wall Street Journal article based on a study by Giovanni Peri and posted it at http://econdataus.com/wsjstem.htm. The study on which it is based is referenced in another paper on the White House web site titled The Economic Effects of Administrative Action on Immigration. Following are two of the findings from the study that the paper repeats:

In a more recent paper, Peri, Shih, and Sparber (2014) find that increases in high-skilled immigration have no effect on the likelihood of unemployment for either college-educated or non-college-educated natives. (page 9)

Peri, Shih, and Sparber (2014) find that increases in high-skilled immigration are associated with significant increases in wages for both college-educated native workers and non- college-educated (high school graduate) native workers. (page 10)

In fact, I emailed Peri and asked him for the data and calculations to replicate this study and he declined saying that he had discussed it with both of his co-authors (Shih and Sparber) who he CC'd on the reply. The reason given was that they were revising the paper for publication in a journal. This underlines the fact that this paper, like the Zavodny paper, was a working paper which, to my knowledge, has not been peer-reviewed, much less replicated.

As I mentioned, I don't expect for anyone to accept my findings as true without careful verification. For that reason, I have posted all the the programs that calculated those findings online. Similarly, I hope that you will not blindly accept the results of any other studies or simply accept a finding because it is repeated in a number of studies. The studies need to be peer-reviewed and/or replicated. Finally, I hope that you will ignore any study for which the author does not release the data and calculations. That would include the Peri paper that I mentioned above.