CS enrollment, foreign graduate students, Gates offshoring comment, etc.

Latest from Matloff...

Date: Sun, 3 Jul 2005 17:17:41 -0700
From: "Norm Matloff" <[email protected]>
Subject: CS enrollment, foreign graduate students, Gates offshoring comment, etc.

To: H-1B/L-1/offshoring e-newsletter

For many of you, the most interesting of the articles enclosed below is
the one in which Bill Gates warns U.S. firms against offshoring their
R&D work. I hope you go beyond that, though, as there are much broader
issues involved, which have allowed me to bring together several
articles from the backlog I have of materials to post and discuss in
this e-newsletter. So, while my commentary here will be a bit long,
please take a little time to go through the whole thing.

As many of you will recall, computer science enrollment at U.S.
universities has been plummeting in recent years. The "official"
explanation of this has been (a) the letdown after the dot-com bust and
(b) the major publicity in the popular press on offshoring of software
development work. To that I would add (c) the increasing tendency of
U.S. employers to fill on-site jobs with H-1Bs.

I would also point out that even (a) is fundamentally an issue of the
shift of U.S. firms to using foreign labor, whether in offshoring or in
imported H-1Bs. The industry's shrill and false claims in the 1990s of
a software labor shortage not only provided the basis for Congress's
expansion of the H-1B program, but also produced a huge, unwarranted
growth in the labor supply, as CS enrollment soared as a response to
the
"shortage" claims. Those hordes of additional graduates hit the job
market just when the dot-com hysteria stopped, causing huge
oversupplies.

As I've reported on a number of occasions, this implosion of CS
enrollment has been causing the CS academic establishment major angst.
To them, this is a horrific threat against their very way of life.
If this reduction in enrollment turns out to be permanent (which I
believe it will), they will suffer the worst fates known to academia:

* Their departments will contract, since faculty size is largely
determined by undergraduate enrollment.

* The amount of research money--of supreme importance--will shrink,
due to having fewer faculty.

* Even with smaller faculties, there will be too many professors
chasing a fixed number of research dollars.
And it has already become very
difficult to get research money, again a consequence of CS
academia's unholy alliance with the industry lobbyists.

* There will be fewer graduate students, due to reductions in both
the domestic undergraduate pipeline and the foreign graduate well.
That means lower PhD production (another supreme issue), and
horror
of horrors, a forced shift from teaching graduate courses to
teaching at the undergraduate level.

I must again make the disclaimer that I am not anti-research. I
consider research to be one of the big attractions of a professor's
job.
But I believe that we should be scholars, with funding being a means of
doing research, rather than an obsessive end in itself, which is what
it
has become.

I've also stressed the point that the CS academic establishment has
itself to blame, because they were complicit with actions taken by the
industry that have now come back to haunt academia: Support of the
industry's false claims in the late 1990s of a tech labor shortage;
support of the industry's push for expansion of the H-1B program; and
support for offshoring. These actions were in the short run a boost to
academia's own expansion, but now have become the very source of CS
academia's current woes:

* As explained above, those actions caused the poor job market
domestic undergraduate enrollment to plummet.

* The poor job market has also caused a big reduction in interest
among foreign students in U.S. graduate study.

* The huge overexpansion in national U.S. CS faculty numbers, a
direct result of the industry lobbyists' shrill "shortage" cries,
now means that there is much less government research funding
available per faculty member, even though overall funding has
increased somewhat.

For my previous postings on these issues, see any of the files with
names beginning with "CS" in my archive of selected postings of this
e-newsletter, at

http://heather.cs.ucdavis.edu/Archive

Now, let's take a look at the enclosed articles in the context I've
outline above. First, the article with the comments by Bill Gates,
calling upon U.S. firms to keep their R&D work here in the U.S.

In this expanded, increasingly competitive economy, India and China
are training engineers who are driving their economies forward, yet
Japan and the U.S. aren't keeping up, he said. At a national level,
both the U.S. and Japan need to train more and better engineers if
their economies are to stay at the cutting edge of technological
innovation, which would create value that helps support both
countries' high standards of living, he said...

"The number of students in engineering and IT is going down. ...
Staying ahead means setting a very high bar," Gates said.

Gates has been eating too much of his own dog food. None of this makes
any sense. It's absurd for Gates and the rest of the industry to
expect
today's young people to go into a field in which those same industry
firms are laying off engineers and offshoring the work. Note
especially
that the July 1 edition of the San Francisco Chronicle had news of a
deal between Microsoft and the Indian offshoring/H-1B giant Tata
Consultancy Services to develop offshoring facilities in China.

Gates' claim that everything will work out as long as R&D is retained
in
the U.S. is unrealistic. First of all, that won't draw in the young
people, because (a) it's too much of a gamble, to go through a CS
major,
first for a Bachelor's and then a Master's degree, on the slim chance
of
getting an R&D position, and (b) many if not most of those "domestic"
R&D jobs will go to H-1Bs, not U.S. citizens and permanent residents.

Now, as I said, the CS academic world is in a real panic over the
prospect of having to live in a world of greatly reduced expectations.
So, they're engaging in a campaign to reverse the enrollment declines,
and in so doing are not being very careful with the facts. One false
claim that is being made repeatedly is exemplified in the letter to the
editor enclosed below, written by an engineering dean at Purdue. It
cites a highly out-of-date BLS projection made on the basis of 2002
data. You see the same BLS study cited in another article enclosed
below, in which a representative from Microsoft bemoans the decline in
female CS students, again citing the BLS study as indicating great
opportunities ahead. I've seen such claims made numerous times in
academia, including at my own university.

These academics may not be deliberately and outrightly lying, but they
certainly aren't questioning what they are passing on the trusting
young
people as firm fact even though the academics have enough sense to know
that the projection is badly out of date.

To put that BLS projection in proper perspective, it forecasts a 45.5%
increase in software engineers between 2002 and 2012. Compare that
with
the 13,000-worker DROP in software engineers during the first quarter
of
this year (during a period of putative recovery in the tech sector),
according to statistics by...the BLS! See the IEEE-USA press release
enclosed below.

This misrepresentation by CS academia for their own personal gain is
really unconscionable. Recently I talked to a retired UCD life
sciences
professor who mentioned that her grandson would be starting as a
first-year student in computer science at UCSD next fall. When I told
her that future job prospects for technical work in this field were
poor, she said that her grandson had been worried about this but the
folks in the Computer Science and Engineering Dept. at UCSD had really
given the kid a high-pressure sales pitch.

Note that the IEEE-USA press release cites BLS as reporting an increase
in jobs for "computer scientists and systems analysts." The term
"computer scientists" is vague, but generally refers to a certain
subset
of R&D positions, an area that even Gates concedes is shrinking in the
U.S. Thus the reported increase is likely in system analysts. What is
going on there?

First of all, as I've mentioned so often, most underpayment of H-1Bs IS
done in full compliance of the law. There are so many loopholes in the
law that there is no reason for employers to use illegal methods. The
system analyst job category has been a favorite refuge for such
employers, as shown dramatically in the Programmers Guild Web page,
"How
to Underpay an H-1B," at

http://www.colosseumbuilders.com/Guild/h1b/howtounderpay.htm

The term "system analyst" is both broad enough to include programmers,
thus
fulfilling the needs of these employers and of a nature that lends
itself to low salaries (again fulfilling the needs of these employers).
That nature is that the term "system analyst" is an old-fashioned term
mainly used for people working on IBM mainframes. Due to the
non-modern
nature of their jobs, their pay tends to be lower. In other words, the
employers of the H-1Bs can hire people with modern skills but pay them
at the same levels as those without modern skills, a typical--and quite
legal--loophole. (The Programmers Guild case study also involves what
I
call Type II salary savings due to hiring H-1Bs.

 

Please read this:

The Economist