Summers Tribute to Kenneth Arrow

Larry Summers is not one of my favorite economists. Maybe I don’t know enough to make judgments, but I can’t forgive him for railroading Bill Clinton into overly relaxing controls on Wall Street in 2000, eventually contributing to disaster in 2008. But what a beautiful remembrance Summers has written about his uncle, Nobel Laureate Kenneth Arrow, who just passed away.

As Summers notes, his other uncle was another legendary economist, Paul Samuelson, “Summers” being an anglicization of “Samuelson.” Summers writes of a family party, held when he was in college, in which the two uncles got into an excited discussion over an arcane mathematical concept, Pontryagin’s Maximum Principle. It was late at night, and the two men’s wives were getting antsy, presumably having had long experience bearing these interminable discussions.

It reminded me of an incident some years ago in Berkeley. I was sitting in a cafe near the campus, and in came famous statisticians Leo Breiman and Peter Bickel, who sat down at a table near me. (I had met Breiman a few times but didn’t know him, and had never met Bickel, and thus was not involved in their conversation, just happened to overhear some.) There is no Nobel Prize in Statistics, but both men would deserve one. (Both are in the National Academy of Sciences.)

Well, they too got into a highly animated conversation about the latest theorem Breiman had proven. He said, excitedly, “It turns out that…,” math talk for “I just proved something quite counterintuitive.” Breiman was near 80 at the time (he passed away not too long afterward) and Bickel was in his 60s. I marveled that they had such enthusiasm for their work, talking like a couple of teenagers on the latest rock group.

Breiman’s passion for statistics (his work is also widely cited in machine learning) manifested itself in other ways, notably K-12 education. Among other things, he once held a seat on the Santa Monica school board.

Good for Breiman and Bickel, and for Summers’ two famous uncles (and OK, good for Summers too). They exemplify what is missing in all the talk of the importance of STEM in the U.S. Most STEM research today is very incremental, almost formulaic, done by people who are obsessed with climbing the academic ladder but with no true passion for their fields. Frankly, a lot of it is make-work, strategically designed to bring in the most grant money and maximize the length of the researcher’s CV. It would be unimaginable for those CV builders to run for the school board, or basically do anything other than robotically accumulating all those increments. And the amount of self-promotion these days is appalling. None of this is going to bring the major breakthroughs that will help the U.S. continue to lead the world in STEM. And don’t get me started on Tiger Mom-ism.

We need more people like Arrow, Samuelson, Breiman, Bickel and so on, who are first and foremost scholars, not academic entrepreneurs. Yet our system more and more discourages that kind of thing, and rewards the CV builders. Sad story.


A Day Without Immigration Rhetoric

I have long complained that The Great Immigration Debate is short on both facts and participants (Congress doesn’t seek our input). So in response to the recent A Day Without Immigrants campaign, I am hereby declaring  A Day Without Immigration Rhetoric. In order to give everyone time to prepare, I’ve set the date to be April 1. 🙂

OK, OK, forgive the tongue-in-cheek lead, but of course I was absolutely serious in that first sentence. The two sides have been talking past each other, typically with unsupported, sometimes deliberately deceptive arguments. Those who know my view won’t be surprised to hear me say that I find the Restrictionist side to be the more responsible one, but only relative to the Let in As Many As Possible side.

What sparked this post today was an article in the California Aggie, the student newspaper at my university, UC Davis (it does not appear to be online yet), which unwittingly brings up a number of issues related to these themes.

The article’s headline and lead report that Ali Baba, a popular Middle Eastern fast food cafe in town, closed in observance of A Day Without Immigrants, especially in protest of President Trump’s recent temporary travel ban; Ali Baba terms the action “unconstitutional.” The article then gives the usual pro-immigration quotes, but I was startled by one from a DC think tank that is generally regarded as leaning toward tightening up a bit in national immigration policy.

That think tank is the Economic Policy Institute. The article refers to a quote by ABC News of EPI Director of Immigration Policy and Research Daniel Costa, who noted that if all immigrants were to disappear tomorrow, our nation’s economy would be devastated. He cited construction and IT as examples.

I have enormous respect for EPI in general and Daniel in particular. But clearly he was setting up a straw man. Any sudden disruption to any system would likely cause turmoil, but that is irrelevant to the question of designing good long-term policy.

Moreover, the quote illustrates one of the most prevalent and pernicious types of fallacies in the immigration debate, the notion that without immigrants from Widgetland, we would have no one available to work in our bustling widget factories. And that is indeed a fallacy in most cases, sometimes directly demonstrably so.

Take construction, for example. Many of you will recall that after the ravaging of the New Orleans area by Hurricane Katrina, some local African-Americans were hired to help rebuild. But after a couple of days, their supervisor told them, “Y’all are fired, because the Mexicans have arrived.” Or consider the case of the security baggage screeners at SFO. They used to be almost exclusively Filipinos, and it was claimed that this simply wasn’t a job that mainstream people wanted to do. But after 9/11, Congress mandated that (a) the workers be citizens (the Filipinos were mainly green card holders) and (b) the workers be paid a decent wage. Lo and behold, TSA was inundated with applications from mainstream Americans. Today, the screeners at SFO are models of — dare I use the term? — diversity.

I teach computer science, and I would estimate that at least 75% of my American (not foreign student) undergraduate students are children of immigrants. You’d be amazed to see how many of them have first names beginning with Z — Chinese, Eastern European, Indian, Middle Eastern and so on.  In other words, without immigration, most of my students wouldn’t be in my classes. And for that matter, without immigration, neither would I myself be in my classes; indeed, I would not exist. 🙂 But I won’t flatter myself into thinking that there wouldn’t be someone else teaching CS instead of me, and that there wouldn’t be a student body whose roots go back more than just one or two generations. After all, most universities worldwide do not rely on children of immigrants to populate their classes.

In fact, in the case of IT, I have written — ironically, in a report for EPI — that the immigrant workers in the CS field are on average slightly weaker than their native peers. Note carefully that my analysis involved the former foreign students in U.S. universities, the immigrant group the industry considers the best of the ones coming from abroad. In other words, I have argued that on the whole, immigration has harmed the tech industry, and with it the American economy and ability to compete internationally. (I have thus called for stringent restrictions on work visas and employer-sponsored green cards,  but liberalizing the aspects of our current policy for visas and green cards for the truly Best and Brightest.)

One more point about the Aggie article: In reporting the widespread participation in A Day Without Immigrants, the story cites rates of nonattendance in various schools that day. In particular, it seems, a KIPP school in Dallas had droves of kids staying home. Fine, but KIPP schools are designed to help disadvantaged, low-income, typically minority kids do well academically and aspire to college. Assuming that a substantial number of those Dallas KIPP kids were from immigrant Latino families, the incident illustrates a sad truth: Immigration policy has winners and losers, and the losers tend to be minority/low-income folks, especially African-Americans. We know from study after study that this group is harmed in terms of work opportunities (see Katrina above), but it is seldom mentioned that resources aimed at improving academics for the lower class are spread thin by the large influx of working-class, poorly-educated immigrants. We must serve ALL kids in this category, but with highly limited resources to do so, immigration has a direct adverse impact.

In other words, the KIPP example, which the article is presenting as evidence that the populace wants more immigration, actually is an argument for having less immigration.

There are many such considerations arguing for tightening up on immigration. At the same time, there are other reasonable arguments for liberalizing policy. Where do we draw the line?

We as a nation need to discuss just what it is that we want from immigration, and how to achieve it. Unfortunately, no one is asking for our input on this, but even if someone did so, we be so handicapped by the paucity of fair, insightful analysis that our inputs would not be so meaningful after all. So, see y’all on April 1.



In Reforming H-1B, Don’t Forget OPT

OPT (“Optional Practical Training”) is part of the F-1 student visa program, allowing foreign students to work in the U.S. for a certain period of time, ostensibly to get practical internship experience before returning to their home country. Last year the Obama administration extended OPT to a 3-year period for STEM students. More on the details below, but first here is the central point of my post this evening.

I was startled today to read the following in the Chronicle of Higher Education (“From China to America. Then What?”,  January 29, 2017):

Last year there were more than 52,000 Chinese graduates on OPT alone. One in every six student-visa holders from China is, in fact, working, not studying, in the United States.

52,000! And 16% of the Chinese students working rather than studying. These numbers are not new, or are at least deducible from published figures, but seeing them described in these terms really took me aback.

And according to the pro-F-1 organization IIE, the overall number is 147,000. Since the number of workers holding the H-1B visa is on the order of several hundred thousand, it is clear that OPT is a major issue.

I’ve emphasized that H-1B is largely about age — employers hire young H-1Bs in lieu of older (35+) U.S. citizens and permanent residents. Since OPT is for new graduates, who almost always are young, OPT has a big impact on older American workers.

The impact is also on the American new graduates. A number of people have pointed out that OPT workers, as “students,” are not subject to FICA tax, thus making foreign students somewhat more attractive to employers as hires than domestic students. But there is a more subtle, and much more harmful impact of OPT as well: A foreign student can use OPT time for an internship, say during a summer, BEFORE she graduates, as Ms. Wu is doing in the Chronicle article. Since internships are hard to get in STEM (the article uses the word coveted), the competition between foreign and domestic students for internships, enabled by OPT, is a very serious issue.

Note carefully that OPT did not arise out of legislation. Instead, the executive branch, many years ago, devised it on their own, declaring a post-graduation internship to be part of being a student. The original duration was one year, but was increased to 29 months by George W. Bush and then 36 months by Obama. As Prof. Ron Hira has pointed out, the idea that someone graduating with a Master’s degree then needs a 3-year “internship” is preposterous.

Not only is the extra-legislative provenance of this policy suspect, but also the original intent has been completely ignored. Supposedly the idea of OPT was to give foreign students practical experience to take back to the impoverished Third World. Yet last year even the DHS admitted that OPT is now used mainly as a holding pattern by employers of F-1 graduates who are waiting to “win” the H-1B lottery.

In other words, OPT is basically a farce — a 147,000-worker farce.

The Chronicle article profile of Wendy Wu, and a type of case I’ve seen many times: The student drifts around in the U.S. from one school to another (3 in Wu’s case), or from one academic field to another, buying time until something comes through that allows her to stay. My aim here is to criticize policy rather than Ms. Wu, and some here know that I am partial to the Cantonese :-). Good for her for her persistence and grit, and I am sure she is a competent worker.

But her LinkedIn profile indicates that she is typical of the phrase I use for the foreign tech workers — ordinary people doing ordinary work, from ordinary schools. The vast majority are not The Best and the Brightest.

Some may object, pointing out that Wu does have an internship at Google, a firm with high standards. But it’s just an internship, not necessarily leading to a permanent job, and frankly, often coming largely from having the right connections. And as I have stated before, Google gives unwarranted preference to hiring foreign workers.

I will take Wu individually to task on one point. She is quoted in the article as follows:

She had come to the country legally, paid her taxes, and followed its laws. It frustrated her to be labeled a job-taker, an interloper. In fact, a recent lawsuit, filed by a group of high-tech workers, accuses foreign graduates in the optional practical training program of snatching opportunities that would have gone to Americans.

Wendy found such anger misplaced. Thousands of technology-sector positions, she says, are going unfilled, and workers from overseas are helping meet the demand. (Whether or not there is a talent drought in the high-tech work force has, unsurprisingly, become embroiled in political debate.)

Besides, she wasn’t looking for a leg up. When, at the end of her first year of graduate school, she won a coveted Google internship, she knew she had earned it, through perseverance, focus, and hard work.

“This is a competitive market. Nothing’s handed to you,” she says. “Stop arguing about where I am. Look at how I got there.”

It is this sense of entitlement that has often bothered me. That 147,000 figure vividly shows that foreign graduates working on OPT do have an adverse impact on new American graduates (not to mention the impact on the older Americans). And of course her claim that there is a tech labor shortage is simply false, as has been shown statistically and otherwise.

The Trump administration is currently working on reform of the alphabet soup of work visas. Hopefully they won’t forget OPT. At a minimum, the Bush/Obama extensions of OPT need to be rolled back.

Identity Politics and Foreign Students from China

UC San Diego has invited the Dalai Lama to be its commencement speaker this year, enraging the Chinese Students and Scholars Association at the campus. CSSA is vigorously protesting, raising some interesting issues.

One tack taken by CSSA is rather novel, invoking “diversity” issues. Having the DL speak would be anti-diversity, they argue. CSSA’s Chinese-language statement seems to imply that UCSD’s action has even offended Chinese-Americans, not just Chinese foreign students. So far, though, invoking Diversity has not been the showstopper CSSA expected it to be.

Particularly interesting is this aspect:

In a statement it posted on WeChat, the Chinese Students and Scholars Association at UCSD said it contacted the Chinese consulate in Los Angeles for guidance and engaged in negotiations with “relevant departments” at the university.

It is interesting in these days of Mike Flynn and related controversies, with the Trump administration accused by some in terms tantamount to treason, that a foreign government — one that all major presidential candidates exhibited some hostility to in 2016 — has close ties with the largest (by far) group of foreign students in the U.S.

Mind you, for the Chinese foreign students, the Chinese consulate is THEIR consulate, representing THEIR government. They have every right to “seek guidance” from the consulate.

But on the other hand, the vast majority of students from China do stay in the U.S. after graduation and become Americans. This has always been a point of ambiguity, ever since the first wave of Chinese students came to the U.S. in the 1980s.

For example, after the violent protests and quashing of same in Beijing in 1989, the Chinese students demanded that Congress give all Chinese foreign students in the U.S. green cards (though, curiously, not political asylum), on the grounds that the Chinese government was spying on them. Yet a few years later, green cards in hand, the Chinese protest a report on CBS News that claimed China was sending a lot of spies to the U.S., many of them students.

The other irony, of course, is that the Chinese students speak of the DL as “wanting to split our country” as if such a desire were evil, at the same time that in California, where these students are living, there is an active Calexit drive aimed at exactly such a split. Not to mention the fact that in the UK, such drives were on the official ballot, first for Scottish independence and then for Brexit.

UCSD may have underestimated the pressure the Chinese government can bring to bear on this issue. Granted, it didn’t work with Obama, but at a university things can be very different, especially with the large number of students and faculty at UCSD who are originally from China. The university’s chancellor will likely stick to his guns, but this will be interesting to watch.

Hatch Talking Trump into Going Lightly on Tightening Up on H-1B?

Thanks to an alert reader of this blog for bringing to my attention the article, “Hatch Doesn’t Expect Trump to Weaken H-1B Visa Program,” in Morning Consult. Some excerpts:

Hatch believes his friendship with Trump — he was one of the first senators, along with Sessions, to endorse his candidacy — means he can convince the president that the data on H-1B visas shows how the program benefits American workers and the U.S. economy.

“You count on me getting that across to him,” Hatch said during Tuesday’s phone interview. “I think I have already, but I’m going to continue until there’s no question he understands me.”

A Hatch aide said last week’s meeting was one of several since the election, and Hatch said there have been other occasions where he presented Trump with data showing how the H-1B visa program promotes industry growth and creates American jobs by filling the high-tech “skills gap.” He pointed to a 2012 study estimating a shortage of more than 220,000 workers in the science, tech, engineering and math fields by 2018…

“While some have expressed some reservations about the impact of high-skilled immigration on American jobs, I believe we can and will be able to make a convincing case for reform,” said Hatch. “The data is on our side.”

Well of course the data is on their side — they paid for it. That study they are referring to is funded by a pro-H-1B advocacy group. I’m told the group pays researchers $50,000 to write a study favorable to the group.

Immigration Lawyer Concedes That the Intels Underpay Their Foreign Workers

Yesterday an NPR piece featured an interview of Bay Area immigration lawyer Ann Cun. Her tone was refreshingly mild, and she made one big concession:

(host) SIEGEL: One criticism of the H-1B visa program – it’s a criticism against guest worker programs – is that they limit the visa holder to working for one employer. So the employee has no bargaining power over pay or promotions, unlike an American who would be free to go work for someone else. Is that a fair criticism?

CUN: I would say initially not necessarily. When you’re in a very competitive market and you’re negotiating with a particular employer, you have the upper hand to negotiate a compensation package that is consistent with the industry and consistent with your peers. Now, on the other hand, if you’re tied to one employer long term for X number of years and you’re relying on that employer to continue to sponsor you for a visa, I could see over a long term period that power of negotiating can diminish over time.

As I pointed out in a recent post, even at the time of hire, the foreign workers have less negotiating power than do U.S. citizens and permanent residents, because the work visa and possible green card are of huge value to them. Indeed, even a pro-H-1B person quoted by the San Francisco Chronicle article I discussed in my post said a foreign worker might be willing to take $25,000 less in salary in exchange for green card sponsorship.

So Cun is being naive here in saying that starting salaries will be equal for equally-qualified Americans and foreigners. But at least she admits that, once hired, the foreign worker’s negotiating power rapidly goes away.

Unfortunately, most of the report has the “Intels Good, Infosyses Bad” theme that I have been saying is so destructive, and that issue matters here in terms of mobility issue Cun brings up. The fact is that immobility is worse under the Intels.

If the employer is simply sponsoring the worker for the H-1B work visa and not a green card, the worker is fairly mobile. He will not be subject to the H-1B lottery or cap if he moves, and the bureaucracy involved is minimal.

But if the worker is also being sponsored by the employer for a green card, the picture changes dramatically. He now would have to begin the green card process all over again if he were to switch employers. Though he may be able to retain his priority date, he still would have to go through labor certification and so on again.  Even more important, there would be lots of uncertainty involved. What if the new employer reneges on his promise to sponsor for a green card, or drags his feet? What if the new labor certification hits a snag? What if the new employer is a startup, with funding in hand for two years but merely hope for what happens afterward?

The Infosyses almost never sponsor their workers for green cards. (See Ron Hira’s work if you need numbers.) Unfortunately, some people use that as an argument supporting the “Intels Good, Infosyses Bad” point of view — the Intels are claimed to use H-1B responsibly, because they sponsor for green cards. But it is exactly the opposite — since the Intels sponsor for green cards, and then use that sponsorship as a way to trap their foreign workers, the Intels’ abuse of the foreign worker system is actually WORSE than that of the Infosyses.

But, you might counter, isn’t it true that the Infosyses generally pay less than the Intels? True but irrelevant. The two sectors are hiring different classes of people. I’ve made a car analogy on this in the past: The Infosyses are buying Toyota Corollas while the Intels are buying Camrys — but both are getting a 20% discount on the classes of cars they are buying.

And once again, the wages are not even the major point. A much bigger issue is loss of job opportunities for Americans.  To suddenly lose one’s career at age 35 or 40 is far worse than having to take somewhat lower wages. Both the Intels and Infosyses are employing foreign workers in jobs that could be filled by Americans.

And in turn, one of the most dangerous aspects of the “Intels Good, Infosyses Bad” mythology is the emphasis on the word replace, as in “The Infosyses are using H-1Bs to replace Americans.” But the Intels hire H-1Bs instead of Americans. There is no difference.

The end of the NPR piece is priceless:

SIEGEL: And we should acknowledge NPR has a small number of H-1B workers. Last year it filed for three applications for H-1B visas with the U.S. Department of Labor.

Yes, H-1Bs are hired in the journalism field too. My favorite example was given to me by a reporter at the Dallas Morning News. Turns out the paper was hiring an H-1B as a “bilingual sports photographer.” No qualified Latino-Americans down there in Texas? Wonders never cease.

Mike Flynn

If I had to bet on it, I would guess that Mike Flynn is toast. Possibly crossing ethical/legal boundaries is one thing, but lying to VP Pence may well turn out to be a much bigger issue.

So I am not here to defend Flynn, but I wonder where the boundaries are for presidential candidates, presidents-elect and their surrogates. Didn’t Obama economic adviser Austan Goolsby tell the Canadian government during the 2008 election campaign not to worry about Obama’s public anti-trade statements? Obama later tried to smooth it over, but it looks like that did happen.

And what about Obama’s visit to Israel, and meeting with Netanyahu, in that same election campaign? What, you think Bibi and Barry just talked about hummus?

Again, the situation with Flynn is probably different for a number of reasons. But the press would have us believe that there is a solid firewall in such situations, and this seems quite misleading.