|
|
Immigration
Last week BusinessWeek reported a new push to get Congress to address the issue of visas for highly skilled workers. The initiative appears to be coming from a lobbying group called CompeteAmerica, a coalition of employers of H-1Bs and related service providers (e.g. American Immigration Lawyers Association).
The coalition members touted in the BusinessWeek article - Oracle, Google, Microsoft, Motorola, Intel, and Hewlett-Packard - appear to be employers who use H-1Bs in the manner the program was originally intended: namely, to bring in the brightest and the best foreign talent to fill positions for which US workers may not be readily available. However, if you dig into some of the umbrella groups on the membership list, you also find the big offshoring firms - WiPro, Infosys, Cognizant - whose use of the visas is more questionable.
It's probably good that the issue of skilled workers is being considered separately from the much larger issue of how to deal with illegal immigration and low-wage workers. My sense was that fighting over amnesty and guest worker provisions dominated that discussion and could potentially have meant a not-so-well-thought-out handling of skilled workers. This is an issue that it's important to get right, and there are lots of problems with the current implementation. There is a good discussion of some of the current system's problems here
Judging from the CompeteAmerica web site, the first issue the group will be pushing is making it easier for skilled workers to get green cards. That seems like a good move; green cards are preferable to H-1Bs in many ways. Why bring smart, skilled people to the US only to send them away a few years later? While such a measure would increase the number of workers competing for high-skill jobs in the US, surely it's better for the native born to compete with those new workers at US wages rather than competing with them at, say, Chinese or Indian wages.
Looks like a lot of the visa problems created after 9/11 may have been fixed. From MSNBC:
Enrollment of international students in U.S. universities could be showing the first signs of recovery after years of weakness following the Sept. 11 terrorist attacks, as the government refined the visa application process and schools try to get ahead in the global competition for foreign talent.
In the 2005-2006 school year, 564,766 international students attended accredited U.S. higher education institutions, according to the most recent report by the Institute of International Education, a nonprofit partly funded by the federal government to track student mobility in and out of U.S. borders.
The number was flat compared to the year before, but it marked the end of a two-year decline first seen in 2003, which raised alarms within academic circles and among education officials.
Because most international students spend years in their programs, the total enrollment number moves slower than the new enrollment number, which was up 8 percent in the fall of 2005.
As the article notes, total enrollments are a lagging indicator, and this is data from 2005, so it's already old to start with.
The article cites two other sources:
A more recent IIE [Institute of International Education] online survey shows the recovery holding up. In the 2006-2007 school year, 52 percent of U.S. campuses reported increases in new international enrollments, and only 20 percent reported declines.
and
Although international graduate school applications began to rise again in 2005, the total number for 2007 was still 27 percent lower than 2003, according to CGS [Council of Graduate Schools].
I'm not familiar with the IIE survey. The CGS numbers are interesting, but from what I have heard from more knowledgeable people, they don't disaggregate their numbers in some critical ways, which renders their data much less informative than it could be, and they won't share their raw data.
I have been working with the NSF's Survey of Graduate Students and Postdoctorates in Science and Engineering this week, and I realized that this data set could easily provide some more timely information than IPEDS. There are questions about numbers of first-year students broken out by citizenship, so it's straightforward to track first-year foreign student enrollments from year to year. After 10 minutes wrestling with WebCASPAR, I get the following numbers for first year graduate enrollments in S&E programs (excluding health fields):
| Year | US | Foreign | Total |
|---|
| 2000 | 61,036 | 33,304 | 94,340 |
| 2001 | 62,421 | 35,691 | 98,112 |
| 2002 | 70,374 | 33,810 | 104,184 |
| 2003 | 76,498 | 31,217 | 107,715 |
| 2004 | 77,442 | 29,138 | 106,580 |
| 2005 | 80,054 | 30,347 | 110,401 |
There has been quite an increase in enrollments by US citizens, but foreign student enrollments are still well below 2001 levels. It's interesting that the numbers of first-years continued to decline for several years after 2002.
So maybe the visa problem is not quite as fixed as one might hope? Or maybe the headaches were so great that people are continuing to shy away from applying for visas even after things have improved?
On the plus side, it looks like things are on the rise again, but at the rate the 2004-2005 increase suggests, it may be another year or so before we return to 2001 levels.
For context, here are the total numbers of S&E graduate students broken out by citizenship:
| Year | US | Foreign | Total |
|---|
| 1996 | 391,095 | 102,984 | 494,079 |
| 1997 | 383,327 | 103,881 | 487,208 |
| 1998 | 378,560 | 107,067 | 485,627 |
| 1999 | 377,802 | 115,454 | 493,256 |
| 2000 | 364,954 | 128,357 | 493,311 |
| 2001 | 368,840 | 140,780 | 509,620 |
| 2002 | 387,532 | 152,885 | 540,417 |
| 2003 | 412,282 | 154,888 | 567,170 |
| 2004 | 424,147 | 150,790 | 574,937 |
| 2005 | 436,530 | 146,696 | 583,226 |
The first year enrollment numbers aren't available before 2000, but the rapid rise in total numbers from 1996-2003 indicate that there must have been pretty rapid growth in first year numbers from 1994-2001 or so. To have killed off that much growth must have taken quite a fiasco in processing visas and in the US's reputation abroad.
Here's an interesting resource: a Department of Labor site that lets you see who is hiring H-1Bs and how much they are paying them. The UI is awful - looks like the DoL could use some better programmers - but if it hurts your eyes too much, you can download all the raw data and write your own. (In fact, I think that would make a good project, and I bet one could get funded to do so)
A few examples. H-1B hires last year:
Universities
- Harvard: 380
- Duke: 313
- Stanford: 364
- University of California system: 1684
(Hires at the above places appear to be largely lab techs and postdocs)
Tech companies
- Google: 350 (in California). Salaries range from $80K to $120K
- Intel: 367 (in Oregon), 682 (in California)
- Apple: 215 (in California)
- (Whoa!) Microsoft: 4100 (in Washington) Salaries start around $75K
Here's an interesting project idea:
- Download the data sets, merge the efile and fax data
- Normalize the employer names so you can do a search for, say, all Microsoft hires in the US
- Match school names to IPEDS codes so you can do things like get breakdowns of H-1B hires as a function of faculty size, etc
- Extract postdoc salaries and post them by institution
- Look at compensation relative to prevailing wages - who's hiring high-end people vs. low-end people?
- Overall, how are academic institutions (who are exempt from caps) using H-1Bs? What kinds of people are they hiring?
I suspect that at least some of these questions have already been answered and that there may already be cleaned-up versions of the data set floating around econ departments. I bet Ron Hira would know about such things.
Google has started a public policy blog. One of their first posts: Google's stance on H-1B visas (pretty simple: "we want more!")
It's an understandable position - Google hires lots of really smart people from all over the world, and current H-1B limits cause them all sorts of headaches. They claim that there were 70 people they wanted to hire last year but were unable to because of the visa caps. There are other effects, too: a friend of mine was hired by Google but had to wait for 8 months to start working there because of H-1B-related delays.
The Googlers I've met who hail from outside the US are all incredibly talented people. It's great for the US that they are here using their talents. The same goes for the people I met back when I worked at Microsoft Research. As James Fallows writes in The Atlantic this month,
The easier America makes it for talented foreigners to work and study there, the richer, more powerful, and more respected America will be. America’s ability to absorb the world’s talent is the crucial advantage no other culture can match—as long as America doesn’t forfeit this advantage with visa rules written mainly out of fear.
The trouble with the H-1B visa program is that not all companies are good corporate citizens in their use of H-1Bs. There are allegations that many companies use the program not to bring in talent unavailable in the US, but rather to drive down wages or facilitate offshoring. The unfortunate result is that legitimate users of the program, like Google and Microsoft Research, get lumped in with firms engaged in questionable practices.
Google is usually pretty creative in their approaches to solving problems. I'm disappointed that they've taken such a conventional stance in this case.
Here's an idea for them: Why not auction off the visas? Instead of a one-time fee, have sponsors commit to a recurring annual charge. Set the number of visas such that the annual fee is on the order of a few thousand dollars (with discounts or exemptions for grad students / postdocs / etc).
An auction would have the benefit of prioritizing things: the more a company needs someone, the more they should be willing to pay. Want to make sure your superstar rocket scientist gets a visa? Bid high. A few thousand dollars a year is a small price to pay for somebody commanding a 6-figure salary. On the other hand, if your goal is just to hire cheap labor, a few thousand dollars could be a significant deterrent.
$5,000 a year from the 300,000+ H-1Bs in the country at any given time would mean a extra $1.5+ billion/year in revenue to the federal government. That money could be used to (a) improve enforcement of protections for domestic workers, (b) to ensure that the visa approval process is efficient and relatively painless for all involved, and (c) to fund educational improvements for domestic students and retraining programs for laid off tech workers.
Reports of the demise of the immigration bill appear to have been greatly exaggerated. Science has a couple of interesting pieces on provisions affecting scientists that I missed in my earlier look at the bill (perhaps they have been added in one of the many, many amendments?), and they are big ones.
From the first article:
Foreign students earning an S&E masters or PhD from a US university would be granted permanent residency. 2/3 of the visas in the diversity lottery would be reserved for people with advanced S&E degrees (the lottery sets aside 50,000 visas for people from countries that do not send large numbers of immigrants to the US)
From the second article:
Foreign students could remain in the US for 2 years after graduation while trying to find a job that will sponsor them for an H-1B. H-1Bs would be granted preferentially based on a point system that looks at such things as whether one has a graduate degree, whether one has employment in S&E, and proficiency in English.
So, in summary, as best I can tell:
If you get an advanced S&E degree from the US, you get to stay. If you get an advanced S&E degree from anywhere else (or if you already have one), you can come work in the US and potentially stay. If you have an advanced S&E degree and are trying to get an H-1B visa, you are either exempt from numerical caps or go to the front of the line, depending on who you work for.
Wow. I think these provisions and the ones described earlier could have potentially huge implications for the global S&E workforce.
Permanent residency for those earning S&E degrees in the US is great - the current practice of bringing in really smart people, letting them gain valuable work experience in tech companies, and then sending them back home seems perverse. I haven't been able to find the provision in the bill (there are 350 amendments at this point, so who knows what's in there), so I'm unclear on when exactly people would qualify. Immediately upon graduation? Or some time later?
The goal of the bill is to make the US more competitive, presumably with China and India, but it's not clear that that will be the long-term effect. Here's why:
If you are in the US, the value of your S&E advanced degree has probably dropped because you will be competing with a lot more people for jobs. I say "probably" because the new entrants will also help create a lot of new opportunities as well. One thing to consider, though: tech companies and the ecosystem that surrounds them employ a lot more people with advanced degrees than just scientists. There are lots of MBAs and JDs running around the tech sector, and nobody is trying to import more of them. Even if the new opportunities for S&Es cancels out some (or all) of the downside of increased competition for jobs, the net effect is likely that the value of an S&E degree relative to an MBA or a JD has fallen. And that means that smart US students will gravitate away from science toward other alternatives. If you are in China or India, the value of an S&E degree just went up because it's a ticket to the US. That means that students will be even more likely to pursue degrees in these subjects. Some fraction of these S&Es will stay in country, and some fraction of those who come to the US on H-1Bs will return. As James Fallows describes in The Atlantic this month, there are some amazing opportunities for entrepreneurs in China.
So, while the US will get a lot of talent from the bill, China and India stand to benefit considerably as well. And it's not at all clear what the balance will be. Regardless, the world stands to gain a whole lot more scientists and engineers.
Immigration policy has been in the news lately, but the talk has been almost entirely about low-skill workers. Buried in the recent Senate compromise immigration bill, though, are some provisions that will affect scientists and engineers.
There hasn't been much talk about the high-skill worker side of the bill, but what little I have read has been negative. From this past Sunday's SF Chronicle: "H-1B Federal Immigration Bill... Everyone agrees it's flawed, but how to fix it?"
ScienceCareers.org asks, Who Speaks for Early-Career Scientists?. The Wall Street Journal reports that even businesses dislike some of the high tech worker provisions (Business Divided As Debate Opens On Immigration).
So just how bad is it? Here's the bill, S. 1348: A bill to provide for comprehensive immigration reform and for other purposes (who names these things?). Keep in mind that this is not the final text by a long shot - there are already 108 proposed amendments - but it gives a glimpse of the kinds of changes that are being contemplated.
As best I can tell, the proposed changes will primarily affect those with bachelors-level S&E degrees. The provisions that most strongly affect those with advanced degrees were put in place years ago; the current bill expands them modestly, but for the most part leaves them qualitatively the same. There is one change, however, that could have unintended long-term consequences for S&E.
First, some background. There is a quota on H-1B visas that varies from year to year. The quotas are set when Congress tinkers with immigration legislation, which is to say infrequently, so it's easy for the levels to get out of whack with what's going on in the economy. The current number is 65,000 per year, down from 195,000 for fiscal year 2003 (notice that the reduction didn't come until about 3 years after the dot-com crash). Now that the tech sector has recovered, visa requests are up again, and the 2007 quota was reached on the first day of the year.
One thing S. 1348 does is to raise the cap from 65,000 to 115,000. It also adds a provision that the cap will automatically increase by 20% the year after the current quota is reached. (I guess this is better than the 90's approach of making up a series of numbers except for the fact that it's a one-way ratchet. Much saner would be something based on wage levels: have the cap rise and fall based on wage levels for skilled workers - that way recessions would trigger automatic reductions.)
So how much does changing the cap affect S&E PhDs? Probably not much. There are already several exemptions that apply to researchers: First, universities, government labs, and nonprofit research organizations are not subject to the cap at all, so the quota is irrelevant in these sectors. Second, as of FY2005 there have been an additional 20,000 slots for people who hold master's or higher degrees from a US university. This second quota was just reached this year (http://www.immigration.com/newsletter1/h1bmasterreach06.pdf)
S. 1348 expands the set of organizations not subject to H-1B caps: all nonprofits would be exempt, not just nonprofit research organizations, and it exempts those who have "been awarded medical specialty certification based on post-doctoral training and experience in the United States." Sure, this will increase the number of PhD H-1Bs, but because of existing exemptions, it doesn't appear that there have been any serious limits on those with PhDs in recent years anyway.
The interesting bit is that the bill broadens the advanced degree exemption to include those with degrees from non-US universities. Here's why: suppose you are a top-notch Indian or Chinese undergraduate who wants to move to the US. Getting a master's or PhD in the US is a great step toward your goal - the degree provides a stepping stone to an H-1B, which in turn can lead to a green card. Suppose you get into a really elite local university, say one of the IITs, but you fare less well in your US admissions process, perhaps because they have difficulty interpreting your credentials or because of language skills. So you have to choose between one of the best in-country universities or a second- (or third-) tier US university for graduate study. Which do you choose? Currently, if your heart is set on ending up in the US, H-1B considerations weigh pretty heavily in favor of the US university. The proposed change, however, removes that advantage and could reverse the balance. So one consequence of S. 1348 might be to strengthen the top Chinese and Indian universities at the expense of lower tier US institutions.
View archives for September 2007.
|
|
