Point System-Based Immigration Policy?

Reportedly the incoming Trump administration is considering a points system to replace current policy on (legal) immigration. Canada and Australia, among others, have policies along these lines.

The basic idea is to bias policy in favor of immigrants who will be most beneficial to the U.S. Seems plausible. What’s not to like? As the son of a skill-less immigrant who barely finished high school, I do continue to hold the romantic notion of the U.S. being a mecca for “your tired, your poor” and so on, and I have always advocated having a policy that consists of a mixture of socioeconomic classes. But this might be impractical today, and in any case, my interest in writing this post concerns how Silicon Valley, and another interest group to be explained later in this post, might view a points system.

The Trump people seem to think that the tech industry would love the plan, which presumably would bring in lots of engineers and scientists. But I’m pretty sure the industry would oppose the idea outright. I say that because they have done so before, when the Comprehensive Immigration Reform bill was brought before Congress in 2007.

Again, what’s not to like, in this case from the point of view of the tech industry? They want highly-skilled immigrants, right? The problem is that Google and Facebook don’t want skilled software engineers; they want YOUNG software engineers. They don’t want the 35-year-old software developer from Georgia, whether he be from Atlanta or Tbilisi. Most of the foreign workers hired by Google and Facebook are young new graduates.

But…if only points could be awarded for being a young new graduate, the industry might be interested. And lo and behold, that’s exactly what Canada is doing.

In other words, Staple a Green Card, the type of program I have argued vigorously against. It would exacerbate the age problem, reduce wages and job opportunities, increase displacement of Americans by foreign workers (albeit ones who become Americans), and above all, bring down the level of insight and innovation of our technical workforce.

There is another interesting pressure group here. Currently a major part of immigration policy is family-based immigration, part of which is known as Fourth Preference. This is the portion of the statute that allows U.S. citizens (typically naturalized ones) to petition for their adult siblings for a green card. This and other aspects form the core driver of what is called chain migration: John immigrates to the U.S., later naturalizes and then petitions for his sister Mary to immigrate. She brings her husband James, who later petitions for his brother Jared and so on.

There has been heavy criticism of chain migration over the years, and people of various immigration-ideological stripes have proposed eliminating the Fourth Preference. But such efforts have been repeatedly thwarted by heavy lobbying by Chinese-American activist groups, who know that the growth in their numbers has come largely from the Fourth Preference. Would they oppose if a point system were proposed today? Possibly not. I’ve heard that at least one of the Chinese groups thinks that the best way to increase the numbers of Chinese immigrants is through H-1B and the like. A solid majority of H-1Bs is Indian, but the second-largest nationality is Chinese

As the vendors at baseball games used to say, “You can’t tell the players without a scorecard!” Identifying the players — and their “positions” — in the immigration game is getting more and more difficult.




DOL Sues Oracle

The Dept. of Labor has filed a complaint against Oracle, charging discrimination on the basis of race, gender and ethnicity. DOL especially alleges that the firm favors Asians in general, and Indians in particular. This is the same agency that charged that Palantir was biased against Asians, as I reported here in September.

Of course, the connection to H-1B, OPT and so on could be major. Oracle does hire a lot of foreign workers, and the general data on H-1B show that most H-1Bs are Indian. It is also worth mentioning another suit against Oracle, in which a former Oracle manager was told to underpay an H-1B because it was “good money for an Indian.”

At that time I wrote about the Palantir case, I discussed some statistical issues, which seem to be carefully addressed in this Oracle case. DOL found that

Specifically, OFCCP found gross disparities in pay [at Oracle] even after controlling for job title, full-time status, exempt status, global career level, job speciality, estimated prior work experience, and company tenure.

When the Mercury News reporter contacted me today for comment on the suit against Oracle, he asked whether I thought the Trump administration will continue to bring suits of this nature against Valley firms. Interesting question, I told him. 🙂 I said I was hopeful that they will.

H-1B Reform Proposal — a Great Start or a Cruel Ruse?

For some time now, various people and organizations have been proposing a simple, effective solution to abuse of the H-1B work visa: Dole them out according to offered salary, highest first. When the visa cap is hit, we stop issuing visas. Reportedly it is under serious consideration by the incoming Trump administration. I myself support this idea. It would not be a full solution, but would be a great start if properly implemented, though of almost no value if badly implemented, a point I will come back to below.

The proposal is appealing for a number of reasons. It is a clean, market-based approach, trivial to implement and would seem to directly benefit the U.S. economy and society: Assuming that the higher-paid workers have more value to add, this proposal is very attractive.

And unlike related executive actions by George W. Bush and Barack Obama that usurped Congress’ legislative authority, this proposal is clearly within the rights of the executive branch to write regulations; the H-1B statute says nothing about the order in which visas are issued, so the executive branch can set that order as it pleases.

And lo and behold, the proposal has wide support, ranging from strong critics of foreign worker programs to supporters of such programs, notably Rep. Zoe Lofgren of Silicon Valley and the IEEE-USA engineering organization.

Well, wait a minute. The fact that Lofgren and the IEEE-USA support the highest-salary-first idea ought to raise some suspicion. As I have pointed out before, the claimed benefit of the policy may be illusory if it is implemented in the manner advocated by Lofgren.

H-1B and green card law require employers to pay the foreign worker something called the prevailing wage, which is the average salary for a given occupation in a given geographical region, and at a given level of experience. Concerning the latter, four levels of experience — proxies for age — are defined. There lies the problem — would a highest-salary-first policy rank wages overall, without regard to experience level, or would the ranking be done separately within each level?

Lofgren has already proposed that the ranking be within-level. If that is the policy that is adopted, we are for the most part BACK TO SQUARE ONE.

How so? As I have often stated — yes, yes, often harped on — the core of H-1B is AGE. Younger workers are cheaper than older ones (age 35+). Employers, both Intels and Infosyses, love H-1B because it expands the pool of younger, thus cheaper, workers to choose from. The vast majority of H-1Bs are classified as Level I or II, and they are hired instead of Americans, or to replace Americans, who are at Level III or IV.

In a case given wide publicity, Walt Disney company replaced OLDER American IT personnel by YOUNGER foreign workers. Sorry for the all-caps format here, but the age issue is almost never mentioned in discussions of the Disney scandal, and yet it was the central issue, the main way Disney saved money by hiring the foreigners.

But it is not just Disney; it is basically all of the mainstream U.S. tech world. I’ve shown before, for instance, job ads by firms such as Intel and Facebook that specify new or recent graduates (NCGs and RCGs, in HR parlance). They ignore the older American applicants and then hire H-1Bs, claiming a “shortage” of applicants.

Note especially the invention of rather bizarre job titles, such as HP’s concocting the title Associate Software Engineer. Indeed, it was clear from the rest of that HP job ad that it was targeting cheap foreign nationals.

The point is this: Even salaries that are well above average in Level I would be much lower than what established Americans make. Disney would still replace Americans by cheaper H-1Bs, and HP would still aim its “creative” job titles at hiring foreigners rather than Americans.

So, if the Trump administration caves to the lobbyists and defines its highest-salary-first policy to mean highest within prevailing wage level, it will be fraudulent “reform.” It will look good, but for the most part just continue the current tragic situation, in which employers employ foreign nationals in preference to equally-qualified Americans. To be sure, it would help a little at the margin, but it will be end of H-1B “reform.”

If the motivation for highest-salary-first is truly that a high wage means a high contribution to the American economy, then the ranking should be done overall, not within experience levels.


Trump Surprises Tech Leaders on H-1B (?)

In a post here more than a year ago, I lauded Donald Trump’s posted platform policy on the H-1B work visa and related issues. But two days later, Trump clarified his position, ruining the effect of his earlier policy completely, so I posted a followup, ending with the statement, “Trump’s a chump.” What did I object to so strongly? Trump now said, in essence, that he wanted to clamp down on the “Infosyses” — the firms that hire H-1Bs and rent them out to other companies — but thought the “Intels” — the firms that hire H-1Bs directly — use the program responsibly. As I have explained many times, this Intels Good, Infosyses Bad view is inaccurate, and will lead to disastrous results, notably a Staple a Green Card to Their Diplomas policy, giving automatic green cards to all STEM foreign students earning degrees in the U.S.

Contrary to what you read in the press (and even to what Trump himself has sometimes said), Trump has been quite consistent on that issue ever since. In particular, he has endorsed Staple, implicitly and explicitly, many times.

So, in advance of Trump’s meeting yesterday with Silicon Valley CEOs, I was expecting him to endorse Staple again. But I certainly wasn’t expecting this account from Recode:

At the top of the gathering (I may not have the order of all the topics exactly right), Microsoft CEO Satya Nadella brought up perhaps the most thorny issue: Immigration and how the government can help tech with things like H-1B visas to keep and bring in more talent. Nadella pointed out that much of the company’s spending on research and development was in the U.S., even if 50 percent of the sales were elsewhere, so that immigration would benefit those here.

Surprisingly to the group, Trump apparently responded favorably, “Let’s fix that,” he said, without a specific promise, and then asked, “What can I do to make it better?”

This is rather bizarre wording, almost sounding like Trump had never even heard of the H-1B visa, and even more industry-friendly than proposing Staple.

If the above account is accurate, it is troubling in ways that go beyond H-1B. Does Trump get confused that easily? Frankly, I in fact doubt that the account is entirely accurate, but it certainly has me scratching my head.


Political Reality 102

I’ve received a number of reader responses, public and private, to my posting yesterday, in which I pointed out that Congress responds to pressure, pure and simple, and does generally NOT respond to well-reasoned arguments offered without pressure. But these readers point out the obvious: If they speak out publicly, they are subject to blacklisting by potential employers. I myself have mentioned this in the past.

In one case, the applicant had a great phone interview and was flown across the country at the employer’s expense for the on-site interview. On-site, he was asked no technical questions to speak of, just friendly chatting. But one of the people who talked to him brought up the fact that the applicant had made a crtiical statement on H-1B to the press. Guess what! He didn’t get the job.

By the way, another piece of political reality that concerned tech workers must keep in mind: If you simply write a letter to your members of Congress, it will only be tallied, For or Against; no one will read it in detail. And the tallier will likely be an intern, even a high school student, and may not have enough background to even get your For/Against status correct. This can especially be likely if your letter is buried in a mountain of letters from the other side, e.g. H-1Bs, immigration lawyers and so on.

My advice — speaking as someone with no personal stake in H-1B — is to do what the lobbyists do. Call, better yet meet with, your members of Congress or their staffs; meet with newspaper editorial boards (did you think those pro-H-1B editorials come from nowhere?); educate journalists and talk show hosts; go to town hall meetings held by politicians; write op-eds and above all, organize, starting with revitalizing the Programmers Guild. And when you do things, know your stuff, the fine details of H-1B issues; otherwise you will be dismissed as not worth talking to.

And yes, there is some risk, but the alternative is possible legislation and regulations coming from the new administration that will make matters even worse than before, even though they sound like an improvement. I’ve been predicting for years that the eventual “reform” would make the problems worse, not better. And my forecasting track record has been pretty good, as Computerworld‘s Pat Thibodeau once pointed out. 🙂

Political Reality 101

Usually I receive countless proposals from people to demonstrate why the H-1B work visa, OPT and so on are bad things. In the last couple of weeks, the number of such contacts has become even larger. The arguments are, for instance, that he workers’ quality is low, their cheap labor reduces tax revenues, the OPTs are not subject to payroll tax so they have an advantage over Americans in seeking work, often the H-1B program is used to facilitate offshoring, a number of immigrants convicted of industrial espionage for a foreign nation (typically China) are former H-1Bs, the privacy of our medical records is in jeopardy and so on.  ALL TRUE — but useless politically.

Congress doesn’t listen to reason. They respond to pressure. So all those people making these good arguments as above, they must become ACTIVIST if they want anything to change, Trump or no Trump. Most people I hear from have never even contacted their congressperson and senators, much less engage in serious, unrelenting activism.

Isn’t this obvious?

Reforming H-1B and Related Visas: What Needs to be Done

Last night’s blog post was titled, “You Can’t Fix It If You Don’t Understand It,” in which I showed that the industry lobbyists are so deft at presenting a misleading view of the H-1B work visa, and policymakers and worker advocates are so uninformed, that effective reform is impossible. In response, one reader suggested that I write a blog post showing, in one convenient document, what needs to be done. Good idea.

Before I begin, please note carefully the centrally important terms. I will use the term “H-1B” to include not only that visa program, but also L-1, OPT, employer-sponsored green cards, and so on. Also, the term “American/U.S.” refers to U.S. citizens and permanent residents. The “Infosyses” are firms that hire H-1Bs and rent them out to other companies, while the “Intels” means firms that hire H-1Bs directly, especially foreign students at U.S. universities, which is a very wide range, including such firms as the Bay Area chain, Bank of the West.

Here we go:

Understanding the problem:

Again, understanding the problem is crucial to developing an effective solution. President-Elect Trump has stated that he will direct AG Sessions to investigate the situation, but again, unless Sessions understands what to look for — and knows how to keep the industry lobbyists at bay — his efforts will be wasted. These points in particular are absolutely key:

  • Hiring H-1Bs instead of Americans is just as harmful as hiring H-1Bs to replace Americans. Isn’t that obvious?
  • The Intels (who hire “instead of”) are thus just as culpable as the Infosyses (who replace).
  • Age plays a CENTRAL role in the hiring of H-1Bs. Younger workers are cheaper (and young H-1Bs are even cheaper than young Americans), so employers hire young H-1Bs in lieu of older Americans.
  • What percentage of hires of H-1Bs, OPTs and so on are genuinely due to a lack of qualified Americans? I believe it is no more than 20%, probably less, even among the Intels.
  • What occupations — IF ANY — have a labor shortage? What does “shortage” even mean? It should NOT mean, “After we automatically filter out all applicants over age 35, we didn’t have enough applicants.”

Suggested executive actions regarding understanding the problem:

The Labor Condition Application (LCA), which employers submit for permission to hire an H-1B, should be extended with some survey questions, such as “State the mean age among STEM workers at your firm,” “State the percentage of new and recent graduates among your STEM hires this year,” “State the percentage of American applicants for STEM positions who were interviewed, and the percentage of those who were extended offers,” “State the percentage increase, if any, in the wage you are paying today in this occupation relative to that of one, two and three years ago,” and so on. The LCA should also have the employer state whether a graduate degree (MS/PhD) is really necessary for the job in question, and if so, why. These answers would not be used as grounds for denying the LCA (but should be subject to perjury laws), just used as data collection relevant to the H-1B program.

In his investigation as directed by Trump, the AG should be especially wary of taking the industry at their word in terms of claiming labor shortages. For instance, the industry lobbyists cite a “shortage” of new computer science graduates, a highly misleading claim, since about 80% of, for example, software engineers do not have a CS degree. And again, often when they say “shortage,” they mean “shortage of young applicants.” The investigators should also be wary of artificial job requirements, such as listing an alphabet soup of computer acronyms that may not really be required. The investigators should make big use of two very valuable resources, the OES data (DOL) and the NACE (National Association of Colleges and Employers) data, both of which show flat wages, counter to shortage claims.

As part of the investigation, the AG should have teams do site visits to HR Departments at a few firms that use a substantial number of H-1Bs, say over 10% of their STEM labor force, in various locales and industry sectors, to determine how/why American job applicants are rejected, especially those over 35. Again, this is for information gathering rather than grounds for denial of visas.

Many H-1Bs, especially those at the Intels, are hired as foreign students from college and university campuses. Currently many U.S. graduate programs have well over half their enrollment as foreign students, in some cases even 90%. This presumably is an unhealthy situation, and the investigation team should particularly note the issues here.

The investigation should consult with the foreign worker advocacy group, Immigration Voice, especially concerning exploitation of foreign workers who are immobile due to waiting for a green card. This  immobility makes foreign workers enormously attractive to employers; a prominent immigration attorney (and former chief architect of Texas Instruments’ immigration policy) even pitches this point to employers on his Web page, urging employers to hire foreign students instead of Americans.

The President should require that every academic institution that offers Master’s/PhD degrees and receives federal research funds submit a yearly report stating the percentages of American students in each STEM graduate program. Those with low percentages should be required to describe what efforts, if any, the institution has made to recruit more American students. These programs should include “professional Master’s degrees,” which seem to be aimed at attracting foreign students as “cash cows.”

The President should direct the Government Accountability Office to study the negative impact on STEM wages due to the large numbers of foreign students who enter the U.S. labor market. The GAO should also quantify the negative impact on older American workers due to this YOUNG foreign influx. A 1989 NSF report forecast that this suppression of wage growth would make graduate study far less attractive to American students; the GAO should quantify this. Note that the foreign influx also keeps down wages of graduate assistants, thus further making graduate study unappealing to American students.

All of this should be carefully taken into account before proposing a Staple a Green Card to Their Diplomas program (i.e. automatic green cards for STEM foreign students), a proprosal that, unfortunately, both Trump and IEEE-USA have endorsed. The GAO should also estimate the further erosion in STEM wage growth due to Staple. While it is true that Staple would remove or ameliorate the immobility problem, it would have a terrible impact on Americans.

The AG’s report must clearly delineate the role of age in the entire H-1B issue. Consider discussions of prevailing wage, for instance. On the one hand, even strong supporters of bringing large numbers of foreign workers to the U.S. (via whatever route), such as Rep. Zoe Lofgren and IEEE-USA lobbyist Bruce Morrison, admit that often H-1B workers are making only half the wage of Americans. True, but what they are not disclosing is that gap is due not only to a lax general definition of prevailing wage but also due to a comparison of young H-1Bs to older Americans.

DHS should make every effort to allow a worker who is waiting for a green card to receive raises and promotions without jeopardizing the green card application.

As I have stated many times, the U.S. should make strong efforts to facilitate the immigration of the genuine “best and brightest.” The criteria for the O-1 work visa, and EB-1 and National Interest Waiver green cards, should be liberalized somewhat.

In conducting the investigation, the AG should watch for undue influence from industry lobbyists, and should make sure that advocates for American workers be closely consulted. (Note that I would count IEEE-USA as industry advocates, not worker advocates.) Similarly, the AG team should discount the work of any academic who has received money from the industry and its allies.

Executive actions for policy change:

As I have shown before, e.g. with Facebook and Intel, often the big tech companies have job ads earmarked for new/recent graduates. This is likely illegal under federal law, and thus arguably grounds for denying the application. A section in the LCA should have the employer attest to not having any so-earmarked jobs, as a requirement for granting permission for the hire.

H-1B using firms with low mean STEM worker age (regardless of American or foreign) should be subject to an automatic investigation by the Dept. of Justice

A number of people, including myself, have suggested over the years that the visas be doled out in order of wage offered, highest first, then second-highest and so on until the cap is reached. I’m pleased to see that IEEE-USA now has endorsed the idea. Nice, but again the age issue is crucial. Currently the DOL method for determining legally-required prevailing wage for an H-1B worker is broken down into four experience levels, proxies for age. My guess is that the industry lobbyists (and, I suspect, IEEE-USA as well) would push for doing the wage ranking within each experience level, rather than overall without regard to experience. Indeed, Rep. Lofgren’s proposal would take that approach. Needless to say, this would largely defeat the purpose of awarding visas by wage ranking. This is a perfect example of how a good proposal can be ruined by plausible but misleading arguments made by lobbyists.

The Optional Practical Training portion of the F-1 student visa should be abolished, or at the very least rolled back to its original 12-month duration. As Ron Hira has pointed out, the notion that a foreign worker with a Master’s or higher needs a 3-year internship is patently absurd. DHS has openly stated that it isn’t using the program for that purpose, but rather as a holding pattern for foreign students waiting for an H-1B visa. The fact is that if the H-1B system were properly reformed, there would be zero wait for the visa, as the demand would never even come close to reaching the cap.

Congressional actions for policy change:

The top priority here should be fixing the badly broken prevailing wage system. I’ve noted above that the four-tier wage system is tantamount to the federal government bestowing its blessing on age discrimination, but the problems go even deeper than that.

Currently prevailing wage is defined to be the average wage (for the given occupation, region and experience level). The use of the average is fundamentally at odds with the industry’s claim that that it is hiring H-1Bs either because they are “the best and the brightest” and/or because they possess rare skill sets. On the open market, both of these qualities command a hefty wage premium; such workers are decidedly NOT paid average wages. So why is the prevailing wage defined in terms of averages? I agree with DPE President Paul Almeida that Congress should revise the definition of prevailing wage to be the 75th percentile for the given occupation and region (and NOT bringing experience level into consideration).

The L-1 intracompany transfer visa needs to have a prevailing wage requirement (and reformed as above), and the current 1-year requirement for time in the company should be increased to 5 years.

Those in line for green cards should be allowed freedom of movement in the labor market as soon as the employer’s petition is filed. Of course, if the petition is eventually denied, the worker would lose her access to the market.

General remarks on reform:

As noted, both the executive and legislative branches must resist the entreaties of lobbyists in the reform process. These lobbyists are the best in the business, highly skilled at making plausible arguments for what actually is very bad policy.

The importance of a coordinated, integrated solution cannot be overemphasized. As with health care policy, we have what I call the “pillow theorem”: Pounding down on one side of a fluffy pillow causes the other side to fluff up. For instance, I say to those who sincerely want to help American workers, “Ignore the age issue at your own peril!”, as it enters into almost any policy proposal made so far. The same comment holds for the Intels Good, Infosyses Bad fallacy. A proposal, such as that made by Rep. Lofgren a years ago, to institute a wage ranking system for H-1B but ALSO adding a Staple program, should raise alarm bells.

We the concerned public should be vigilant too. We should insist that the investigation be fully public, including all the data, and that there are people whom the Ordinary Joe can contact to give input. We must make sure we are fully informed of the structural problems in H-1B (see above!). And if Jeff Sessions, in presenting the results of his investigation to a press conference, uses the word “replace” within the first 30 seconds, we should collectively groan.