Wednesday, August 25, 2004

previous entry | main | next entry | TrackBack (4)


Offshoring creates jobs in California

Yesterday, Virginia Postrel posted and linked to several stories about a Public Policy Institute of California study on the effect of offshore outsourcing on the Californian economy. Postrel wrote, "The study found that outsourcing actually increases employment in California. Now the Assembly is sitting on the study."

The Assembly may have sat on the study, but it now appears to be available to the public. I clicked over to the PPIC web site and found the report by Jon Haveman and Howard Shatz, which is dated today. Some of their analysis sounds awfully familiar. The good parts (from p. 22-24):

[T]here is evidence that some California jobs eliminated by offshoring are similar to those likely to be created in the state by offshoring. Offshoring can allow the economy to reallocate labor and capital from one set of tasks to another set of higher-level tasks, and California has a strong supply of highly skilled workers who can take on these higher-level tasks. A second mitigating factor for the U.S. economy, and especially for California, is that there is growing world demand for the U.S. provision of services that are similar to those being sent offshore. This is a source of job creation for workers dislocated because of offshoring. California is a significant producer and exporter of these services, and this fact should ease the transition for affected workers in the state.

These two points suggest that jobs are being created in industries and occupations that are relatively similar to those being eliminated. For example, computer programming is one occupation that is projected to be negatively affected by offshoring, and evidence suggests that this is in fact the case. Between 1999 and 2002, 71,000 computer programmer jobs were eliminated, 23,000 of which were in California. Note, however, that offshoring is only responsible for a small fraction of these lost jobs; the technology bust explains most of them.

At the same time that opportunities for computer programmers were declining, more than 115,000 software engineering jobs were created, a disproportionately large share of which – 24,000 – were in California. The transition between these occupation categories – programmers and software engineers – may be unsettling but also may be easier than the transition between many other jobs, and the software engineering jobs pay over $10,000 more per year on average. It is certainly true that not all the workers who held computer programmer jobs became software engineers, and tracing how any such transition might have taken place is difficult. However, the decline of programmers and the rise of software engineers illustrates the fact that as old opportunities disappear – through economic cycles, technological change, and offshoring – the U.S. economy has the capacity to create better opportunities, and in this case, the skill sets required are similar.

As a result, the economy’s adjustment to this new phenomenon need not be as difficult as it was to the shift from manufacturing to service sectors in the 1980s and 1990s. Many of the workers displaced by offshoring have significant skills, and this bodes well for their future employment prospects. This observation is consistent with the evidence suggesting that workers with more skills have less difficult transitions to new jobs; the transitions are faster and involve less wage erosion. This is not to minimize the effect on workers who lose their jobs, especially less-skilled workers, but merely to put it in perspective....

An additional item is worthy of note. It is likely that “outstating” – outsourcing to another state – is a much more important phenomenon than is offshoring for California. The recent mass layoffs report of the Bureau of Labor Statistics noted that of the job relocations where the destination was known, more than 68 percent took place within the United States, rather than overseas. California companies have actively engaged in outstating back-office processes for years. These are the very same processes that are the most vulnerable to offshoring. It is possible that many jobs being moved offshore by California companies would have left the state anyway. Newspaper accounts in places such as Phoenix express the fear that back-office jobs that have only recently arrived from California are headed abroad.

I look forward to the California state legislature's efforts to impose a tariff on services from Arizona.

Here's the report's conclusion regarding the bills designed to block the offshore outsourcing of government contracts (from p. 31):

In the end, the policy of restricting government contractors to vendors employing only domestic labor falls short of the optimal choice in several respects. First, on a per worker basis, it is likely to be more expensive than other options. Second, it is likely to assist a very small subset of workers displaced by offshoring. Third, policies banning offshoring are most likely to assist relatively skilled workers with high earnings capacity. In this time of tight budgets, more cost-effective means of assistance are available and should be investigated.

Red Herring has more on the California situation. Daniel Weintraub concludes in the Sacramento Bee:

The bottom line, though the researchers don't put it this bluntly, is that politicians, either from ignorance or malevolence, are trying to scare Californians into believing that offshoring is bad for the economy, and bad for them. The reality is that the opposite is true, and that the proposals seeking to freeze the economy in place will do far more harm than good.

Indeed.

[Sure, that's California. The rest of the country is losing jobs, right?--ed. Not according to this Business Week story from earlier this month]:

Foreign investment for setting up U.S. subsidiaries and plants doubled, to $82 billion, between 2002 and 2003, according to the Commerce Dept. That means 400,000 new jobs, most of them tech-related, figures the Organization for International Investment, a trade association based in Washington, D.C. Over the same period, outsourcing has taken away about 300,000 U.S. jobs, according to tech consultancy Forrester Research. So, on a net basis, foreign outfits have actually added some 100,000 U.S. jobs.

There's plenty of incentive to keep the trend going. For starters, foreign companies often find that having a U.S. base can be a big help when selling to the lucrative U.S. market. That's one reason Fremont (Calif.)-based Infosys Consulting, a subsidiary of Indian outsourcer Infosys, plans to hire about 500 consultants -- most of them Americans -- over the next two years, says Basab Pradham, senior vice-president and head of worldwide sales....

As U.S. companies begin to outsource such mission-critical functions as human resources and finance, they still want to be able to coordinate and oversee such work more closely. "What we're beginning to witness is a change in the [offshoring] business model," says Wipro's [corporate vice-president of human resources Pratik] Kumar. "A lot of outsourcing companies used to be completely offshore. But as they've begun to handle more complex work, they find that they need to have more local expertise deployed."

UPDATE: Ashish Hanwadikar has more links on this.

posted by Dan on 08.25.04 at 04:53 PM




Comments:

"The transition between these occupation categories – programmers and software engineers "

THERE IS *NO* *FUCKING* *DIFFERENCE* BETWEEN THE TWO CATEGORIES.

The title depends solely on the company's whim. It has little or nothing to do with job responsibilities.

A "programmer/analyst" at a bank can move to a job as a "software engineer" at a web firm, without any real change in duties performed. All that changed is that the bank calls its developers "programmer/analysts" and the web firm calls its developers "software engineers".

I hate it when dipshits who don't understand the IT industry try to make points about it.

posted by: Jon H on 08.25.04 at 04:53 PM [permalink]



Jon H: Take it up with the Bureau of Labor Statistics -- it's their distinction.

posted by: Dan Drezner on 08.25.04 at 04:53 PM [permalink]



that's great stuff. i plan to read the whole thing tonight.

also I noticed we're not on your favorite blogs roll. whu? i guess it proves even the best of us makes mistakes sometimes. lol....


posted by: jason on 08.25.04 at 04:53 PM [permalink]



The report was not on the website yesterday, when the rest of us were writing about this story. You had to be a California journalist with leaky sources to get it.

posted by: Virginia Postrel on 08.25.04 at 04:53 PM [permalink]



“[Sure, that's California. The rest of the country is losing jobs, right?--ed. Not according to this Business Week story from earlier this month]:”

Oh my goodness, somebody should inform John Kerry’s campaign. He’s obviously unaware of this news story. I’m sure that we can take it for granted that both Kerry and John Edwards will immediately cease attacking those Benedict Arnold executives. Am I right?

posted by: David Thomson on 08.25.04 at 04:53 PM [permalink]



I don't know why Jon H is so upset. In practice he's right, both titles get applied to the same basic job. But there is a subtle difference. If someone tells me they're a software engineer, then I expect them to know several programming languages. I also expect them to be comfortable with a wide range of problems, solutions, etc.

But Jon H correctly pointed out, the titles can be applied in any way a company desires. And there's the problem that in much of the business "engineer" has more cachet than "programmer". Thus everyone puts "engineer" on their resume/job application/Labor Bureau survey form.

But I think this is a non-issue. My own anecdotal experience agrees with Haveman's and Shatz' report: the grunt entry-level programming jobs, where someone hands you the design and you just code all day, are being outsourced. Meanwhile many of the people who held those grunt jobs are now doing the (usually) higher-paying design and systems work.

posted by: Lance Peterson on 08.25.04 at 04:53 PM [permalink]



Lance writes: "Meanwhile many of the people who held those grunt jobs are now doing the (usually) higher-paying design and systems work. "

So where are they going to come from in the future, if the grunt jobs are being done overseas?

Where's the farm team of design and systems people?

posted by: Jon H on 08.25.04 at 04:53 PM [permalink]



Perhaps "programmer" refers to the person with a high school degree and some trade school training who is programming in COBOL on mainframes at places like Wells Fargo Bank.

Yes, as a practical matter, programmer=software engineer. "Software engineering" is something that those who have been trying to push certification for the last 25 years have used to try to add mystery to the field.

Sadly, things like object-oriented analysis and design (and structured analysis and design, before them) have never provided the hoped-for lift. "Extreme Programming" (XP) was a reaction to the abuses of "software engineering" (ala Rational, et al).

We are back to where we started: get some good people, who have a reputation for producing code that works. Give them access to the people that know what the software needs to do, and try to minimize disruptions. At the end, you have a good chance of producing something that "works". Hopefully, they will have adopted the testing component of XP, and your chances are better.

I was on a project a decade ago that was following strict DoD software engineering processes. The trouble is, they had too many people who didn't know what they were doing. Several of the people who did know left, including me, and the project eventually failed. So much for the power of processes.

Regards,

Jim Bender


http://anglo-dutch-wars.blogspot.com/

http://17th-centurynavwargaming.blogspot.com/

http://kentishknock.com/

http://anglodutchwarsblog.com/

http://dreadnought-cruisers.blogspot.com/

posted by: Jim Bender on 08.25.04 at 04:53 PM [permalink]



The difference between the two terms is only slightly more than the difference between a "housewife" and a "domestic engineer."

What they're probably trying to say is that people who just crunch out code were sent overseas, while the designers were not.

The problem is that eventually those overseas will learn how to design.

It'd be interesting to see the percentage of new technologies that were sent overseas vs. kept here. If we're sending experience with leading-edge technology overseas, that way lies Saudi Arabia.

posted by: The Lonewacko Blog on 08.25.04 at 04:53 PM [permalink]



"A "programmer/analyst" at a bank can move to a job as a "software engineer" at a web firm"

Yeah, but no. I'm a software engineer. I have a degree in computer science and engineering. I studied compilers, databases, microprocessor design, etc. I engineer software systems and products. It's a far cry from a COBOL or JAVA programmer at a bank that took a training course.

posted by: David Mooney on 08.25.04 at 04:53 PM [permalink]



Despite the projected bright future for software engineers, computer science enrollments are plunging - even at MIT:

"At the Massachusetts Institute of Technology, as in other schools across the country, computer science enrollments are dropping, raising questions about the country's future tech leadership.


This fall, there are just under 200 new undergraduate majors in MIT's electrical engineering and computer science department, down from about 240 last year and roughly 385 three years ago."

Source: CNET News
Students saying no to computer science

http://digitalcity.com.com/Students+saying+no+to+computer+science/2100-1022_3-5306096.html



posted by: bhaim on 08.25.04 at 04:53 PM [permalink]



bhaim, do youu think if Dan had bothered to wander over to the CS department at U of C, he'd still post this? Dan, next time I walk into Fry's, I'll let the people know that hte job market is actually up!

posted by: Jor on 08.25.04 at 04:53 PM [permalink]



This analysis is bull. The problem is you are comparing apples (managers) to oranges (workers = engineers or programmers). In any credibly run organization there are more managers than workers. What we are doing with offshoring is sending the workers offshore and keeping the managers at home. Even if the number of managers doubled or tripled, the total number of jobs in the field would be MUCH less, perhaps as much as a factor of ten. Easily reduced by two thirds. Recall that in this field the managers and workers need (need!) the same training.At least the computer science degree is required, maybe more. So what we've done is to offshore the entry level positions.

We're not offshoring only jobs, but our technological future.

posted by: camille roy on 08.25.04 at 04:53 PM [permalink]



I'm a software engineer. I have a degree in computer science and engineering.

No, you're a programmer with a degree in computer science and engineering who calls himself a software engineer. Let me know when you have to, say, pass a state exam before calling yourself a software engineer.

bhaim is right. I mean, you could make money selling a kidney, but it's not exactly a good thing long-term, you know?

posted by: The Lonewacko Blog on 08.25.04 at 04:53 PM [permalink]



Here's the link from above.

Students saying no to computer science

This fall, there are just under 200 new undergraduate majors in MIT's electrical engineering and computer science department, down from about 240 last year and roughly 385 three years ago... The Rutgers University computer science department has canceled some course sections and expects total enrollment in classes in the major this year to be thousands less than its peak of 6,500 several years ago. Saul Levy, chair of the undergraduate computer science program, said the ongoing decline stems from the way students perceive career prospects... "They don't believe in the job market in computers anymore," Levy said.

I also scanned the underlying report, which is available here.

Nowhere in that report does it mention anything long-term. (At least that I could see). It doesn't address the problem of technological innovation being sent offshore. It doesn't address the problems described above at MIT and Rutgers.

Many of the wonderful new programs proposed by our "leaders" - whether in offshoring or immigration - tend to have a whiff of Saudi Arabia or Latin America about them. I guess I should just top up my glass of Kato-Aid and sit on my veranda.

posted by: The Lonewacko Blog on 08.25.04 at 04:53 PM [permalink]



Ok, this is too rich

There's plenty of incentive to keep the trend going. For starters, foreign companies often find that having a U.S. base can be a big help when selling to the lucrative U.S. market. That's one reason Fremont (Calif.)-based Infosys Consulting, a subsidiary of Indian outsourcer Infosys, plans to hire about 500 consultants -- most of them Americans -- over the next two years, says Basab Pradham, senior vice-president and head of worldwide sales....

Can someone please follow the logical consequence of hiring 400 people in America to help you gain more out-sourcing work in America really means?

posted by: Jor on 08.25.04 at 04:53 PM [permalink]



David Mooney writes:

"Yeah, but no. I'm a software engineer. I have a degree in computer science and engineering. I studied compilers, databases, microprocessor design, etc. I engineer software systems and products. It's a far cry from a COBOL or JAVA programmer at a bank that took a training course."

But yeah, I was speaking from direct experience.

I spent 3 years working on a Fed Funds/Eurodollar trading system at a bank. As a "programmer/analyst". Then I got a new job at a dot-com, as a "software engineer". No particular change of responsibilities.

Titles are an artifact of company culture, and little more.

posted by: Jon H on 08.25.04 at 04:53 PM [permalink]



The problem with the Business Week quote is that it does not make clear that foreign direct investment may involve 400,000 jobs it does not necessarilly mean that 400,000 new jobs were created. If the foreign direct investment is just to buy an existing firm it does not usually mean that new jobs are created. But you can not tell from the direct investment data quoted if this is true or not in this case. But the general case is that the bulk of direct foreing investmnt in the US is to buy exiting firm and does not produce the many new jobs as the article implies.

So your comparing the 400,000 jobs in firms bought by foreign firms to the jobs lost to offshoring is not a valid comparison of apples to apples. You are adding craps and apples to get crapapples.

posted by: spencer on 08.25.04 at 04:53 PM [permalink]



Am I missing something in this whole debate?

It seems clear that we are seeing increased trade and competition. The number of jobs that can be done 1000's of miles away has increased due to technology and decreased barriers to capital flows. This means that jobs that once had to be done in the US, lets say by a guy sitting next to a giant computer that lived in the back room of a bank, can now be done without sitting in that back room of the bank. Thus the person who used to have the job sitting in the backroom in the bank now has to compete with someone who has the same skills in Arizona or India.

More competition means a lot of things. It increases the supply of labor which, all things being equal, would push down the price paid for that labor. This has to be the dominant dynamic here. And when the price of that labor falls that means that the price of the service provided, in my example banking, also would fall.

But there are all kinds of secondary things going on as well, and that is I guess where all the debate comes from. You have the person in India who is making more money, consuming more goods and services, some of which are produced in the US. You have more jobs for people who are making it possible for the original job to be done by people in India (but if businesses are making smart decisions the net payments for this should be less than the savings from the increased labor competition).

You also have the guy who was sitting in the backroom in the bank now out there looking for a job and looking for new things that he can do. For some this means they are reading Dan's blog getting angry at him for not caring about the plight of the programmer. But you have other programmers sitting around being creative and finding new ways to put their skills to use. The theoretical view says that eventually those people are going to create new markets, new jobs, and increase prosperity. The personal view says, f-that I want my paycheck and I want to know that my family are not going to be evicted.

I guess the conflict is inevitable, but I still don't understand why each side seems to view this as an argument rather than an effort to understand the very real changes that are going on in our economy.

posted by: Rich on 08.25.04 at 04:53 PM [permalink]



Rich writes: "The theoretical view says that eventually those people are going to create new markets, new jobs, and increase prosperity."

But long term, those people are going to be the ones in India, not the US, because people in the US will be gravitating towards the dead-end service jobs that are available.

And sure, some of the service industry people will create new businesses such as day spas or gourmet pet bakeries, but those industries don't a global powerhouse make.

The technological innovation will move overseas along with the jobs.

posted by: Jon H on 08.25.04 at 04:53 PM [permalink]



Jon H:

OK. You have identified a problem. Grunt jobs are outsourced to cheaper markets. The grunt people get smarter, move up the food chain, so more jobs are outsourced there, or they start their own companies.

With something like software, how does government prevent this from happening, without inflicting worse unintended consequences all over the economy?

posted by: Appalled Moderate on 08.25.04 at 04:53 PM [permalink]



From Salary.com, definitions of "programmer" (level I) and "software engineer" (level I):

Programmer I (IT -- Computers, Software)

Reviews, analyzes, and modifies programming systems including encoding, testing, debugging and documenting programs. May require an associate's degree in a related area and 0-3 years of experience in the field or in a related area. Has knowledge of commonly-used concepts, practices, and procedures within a particular field. Relies on instructions and pre-established guidelines to perform the functions of the job. Works under immediate supervision. Primary job functions do not typically require exercising independent judgment. Typically reports to a project leader or manager.

Median salary in San Jose, CA: $56,693

Software Engineer I (IT -- Computers, Software)

Designs, modifies, develops, writes and implements software programming applications. Supports and/or installs software applications/operating systems. Participates in the testing process through test review and analysis, test witnessing and certification of software. Requires a bachelor's degree in a related area and 0-2 years of experience in the field or in a related area. Has knowledge of commonly-used concepts, practices, and procedures within a particular field. Relies on instructions and pre-established guidelines to perform the functions of the job. Works under immediate supervision. Primary job functions do not typically require exercising independent judgment. Typically reports to a manager.

Median salary in San Jose, CA: $62,271

There are four levels of "programmer", five levels of "software engineer". There are also other, more specialized categories like "Mainframe programmer" and "Client/server programmer", all with four or five levels of experience.

Even though there are separate categories, it should be clear from the description that they are very similar and often, just as Jon H said, indistinguishable. The statement in the report that "programmer" jobs were lost while "software engineer" jobs were created reflects indeed more a trend towards fancier titles than towards more demanding jobs.

The point that there are not enough Americans studying computer science is a very valid one. All the concern about outsourcing programming jobs rings a bit hollow when one considers that so many programmers (or software engineers) employed today are foreigners, hundreds of thousands of them on temporary work visas which can only be issued if a plausible case can be made that no Americans are available to do the job.

I think there are considerably more complicated mechanisms at play here than "outsourcing destroys jobs" or "outsourcing creates jobs".

posted by: gw on 08.25.04 at 04:53 PM [permalink]



"hundreds of thousands of them on temporary work visas which can only be issued if a plausible case can be made that no Americans are available to do the job."

This is actually easy to get around, and companies regularly do so.

posted by: Jon H on 08.25.04 at 04:53 PM [permalink]



This is actually easy to get around, and companies regularly do so.

Well, but there is also the "prevailing wage" requirement, which forces companies to pay the H-1B visa holders the same wage they would pay an American available for the same job.

Add in the visa fees (mostly legal fees), the uncertainty of actually getting the visa issued, the waiting time and last not least the likely communication problems with people whose English isn't all that great, and it doesn't seem likely that too many companies would get H-1Bs unless they really did have trouble finding enough qualified people here.

At the height of the dot com boom I have seen H-1Bs being hired at amazing salaries. In fact, I once pretty much got a raise because an H-1B guy came in at a salary higher than mine and the VP agreed that this wasn't fair. :-)

I have personally been involved in hiring software engineers in 2000, and applications from Americans were few and far in between the many applications from Indians and Chinese folks that we received.

posted by: gw on 08.25.04 at 04:53 PM [permalink]



gw writes: "Well, but there is also the "prevailing wage" requirement, which forces companies to pay the H-1B visa holders the same wage they would pay an American available for the same job."

Well, actually, they only have to pay what someone with the same vague job description would get. They don't need to pay the prevailing wage for someone with the specific skill desired.

ie, if they want a Java programmer, and Java programmers are getting a premium in the market, they can get an H1B Java guy and pay the prevailing wage for generic programmers, thus avoiding the premium for Java.

Also, this: "the new DOL regulation implemented in 1998 sets up two only categories for prevailing wage, Entry Level and Experienced, asserting that this was unfair since the worker with five years of experience will be measured against a prevailing wage calculated on a group that includes people with 25 years of experience."

A good source on H1Bs and L1 visas is Norman Matloff's work at http://heather.cs.ucdavis.edu/itaa.real.html#tth_sEc9.2.5

That was written in the environment when businesses were claiming a tech worker shortage and clamoring for more H1B workers.

posted by: Jon H on 08.25.04 at 04:53 PM [permalink]



Getting back on topic, offshoring and outsourcing are nto the same thing. Most of the outsourcing that takes place today is to US-based firms, and most of this work is simply transferring paper-intensive back-office functions to specialist providers. It's no more controversial or pernicious than externalizing any other non-core cost center function.

As to the pissing match between the "engineers" and the "programmers", the fact is that most entry-level IT jobs require little real skill, and such skills as are required very quickly become obsolete. Even if there were no offshoring of programming work at all, there would still be downward pressure on US programmers' wages because programming skill is a depreciating asset.

Perhaps there's a national security argument to be made for preserving what one poster called "bench strength" in programmers, but I doubt we need anywhere near the number of programmers that the market demanded four years ago when a flood of capital distorted demand for such skills.

Yes, we need enough programmers to maintain military superiority and build the next Google. No, we don't need another 100,000 programmers to create the next Pets.com and stitch together the 5,399th application connector for General Motors.

How many is enough? I suspect the market will tell us. If DARPA can't find enough programmers for its needs, I suspect that the labor price mechanism will respond quickly enough, and students at MIT, Carnegie-Mellon CalTech etc will also get the message.

posted by: lex on 08.25.04 at 04:53 PM [permalink]



This debatre reminds me of the bogus arguments advanced in the 1980s that equated Asian manufacturers' market share gains with a loss of US military superiority. The high (or low) point of this hysteria was reached in Michael Crichton's 1989 camp thriller, "Rising Sun."

Michael Kinsley did a cameo in which he asks a (corrupt) US Senator on TV whether we're selling Japan the technological rope to hang ourselves with. Also loved the Japanese zakuza playboy's karaoke singing of "Don't Fence Me In" and harvey Keitel as a Brooklyn cop in LA (don't ask) who shouts "Gerrrrrrronimo!" before busting into a party held by a US software company that's selling itself to a Japanese conglomerate.

wisten to the murmur of the cottonwood twees...

posted by: lex on 08.25.04 at 04:53 PM [permalink]



lex writes: "How many is enough? I suspect the market will tell us."

If they weren't needed, companies wouldn't be going to India to get them.

Even $20,000 for an Indian programmer is a big expense for something that isn't needed.

Work is being offshored to cheaper locales because programming has proven highly resistant to automation or other process improvements which businesses thought would reduce their need for headcount.

posted by: Jon H on 08.25.04 at 04:53 PM [permalink]



Jon H, from the link you referred to:

A September 28, 2000 article in the Chicago Sun-Times said it succinctly:

``If you're willing to pay market rate, you can find people,'' said Pete Georgiadis, founder and CEO of eBlast Ventures, a company that funds and builds technology firms. ``The issue is if you're budget-constrained, you can't get the people you want.''

True, but highly misleading. If you were willing to pay top dollars and promise a bunch of (ultimately worthless) stock options, you could in fact lure people out of their existing jobs in 2000 and get them to join your start-up.

But it's just not true that there was a large pool of unemployed American programmers standing by in 2000 waiting to be paid "market rate" while Indians were snapping up jobs at 30 % below market. My company at the time paid relatively inexperienced, inarticulate and - as we found out after a few weeks - incompetent Indian H-1B programmers $90k+ because we couldn't find any Americans (or permanent residents!) to do the job.

No, we didn't hire every single American who applied. In some cases we made offers, but were turned down because they had already accepted one of their other five offers. I remember one case where we liked the person, but he was being paid contractor rates and was asking for a full-time salary of $150k. Believe me, that was not the prevailing wage for the kind of job we were hiring for. And he could get it (as a contractor) elsewhere, he was not desperate to be hired.

Also, quite a few of the Americans we did hire ultimately didn't work out very well. The only one who was really good was one of the first to leave because he could easily get an even better offer (see above - lured away). He also happened to be from Hong Kong, but he had become a citizen a while ago.

But a more interesting aspect is that from a free market "outsourcing is wonderful" point of view, the prevailing wage requirement for H-1Bs should really be done away with. Why don't we just let as many people into the country as companies would like to hire and let them negotiate their own salaries with no minimums in place? What's fundamentally/theoretically wrong with that idea? Why aren't outsourcing proponents pushing for that?

Ah, some of our most outspoken outsourcing proponents don't like immigration nearly as much, do they? Don't free markets imply free movement of labor forces across borders? I think I'm sensing some hypocrisy there...

posted by: gw on 08.25.04 at 04:53 PM [permalink]



Sure the skills are needed, Jon H. Just not at anywhere near the price that you think they're worth. Again, programming skill is a depreciating asset.

It's no more reasonable demand that the nation subsidize programming jobs at 3x the market wage than to tell a car buyer he should pay $30,000 for a three year-old Chevy Malibu.

To continue the analogy, if your offering's become commoditized, you should figure out a way to differentiate yourself and add superior value, hence earn a significant market premium, or else leave the market.

posted by: lex on 08.25.04 at 04:53 PM [permalink]



"Why don't we just let as many people into the country as companies would like to hire and let them negotiate their own salaries with no minimums in place? What's fundamentally/theoretically wrong with that idea?"

Because they'd have little or no bargaining power, and end up being paid ridiculously low wages.

posted by: Jon H on 08.25.04 at 04:53 PM [permalink]



gw,

I wholeheartedly endorse the notion of doing away with all immigration curbs (except security-related ones) for anyone with a college degree who wishes to enter this country.

By all means, bring in programmers from Asia, Russia, israel, the arab world Africa, wherever. The more the better, both for us and for their home countries. Eventually they'll return home and transfer not only technology but also, in all likelihood, a favorable view of US capitalism and US democracy.

posted by: lex on 08.25.04 at 04:53 PM [permalink]



Jon H: Because they'd have little or no bargaining power, and end up being paid ridiculously low wages.

Well, yes, but then the same problem exists with outsourcing. Why are we artificially restricing one, but not the other?

lex: Eventually they'll return home and transfer not only technology but also, in all likelihood, a favorable view of US capitalism and US democracy.

Actually, one argument that could be made against this is that many of them won't return home and that we are actually taking away the elites of poor countries, thus leaving them with less talent and even less chance to catch up.

Appalled Moderate asked a good question above that nobody has replied to. I personally don't think that government should actively restrict outsourcing/offshoring. (A different question is whether government should itself engage in it. A lot of the real-life debate focuses on that.)

I do think one should consider measures that would make offshoring for the specific purpose of replacing and laying off people less attractive. Companies who do this could be made to pay for training for the laid off people to help them get into those higher-paid higher-skill jobs we keep reading about. It's all about sharing the benefits and sharing the costs. One could also require companies to provide a certain number of entry-level positions in relation to the number of jobs they sent offshore.

posted by: gw on 08.25.04 at 04:53 PM [permalink]



"It's no more reasonable demand that the nation subsidize programming jobs at 3x the market wage than to tell a car buyer he should pay $30,000 for a three year-old Chevy Malibu."

Ah but you avoid the point.

The job of our political & economic system is to satisfy voters. Right now (perhaps you think this should change?) voting is not restricted to those who are so affluent they have no need to work. That means our political and economic system (gasp!) is not geared to producing corporate profit above all else. If the system has been gamed in that respect, that only occured because general prosperity allowed both the middle class and the corporate sector to flourish.

Offshoring changes this equation. The American worker is being abandoned by American business. Yes this trend started with manufacturing. Yes it has been extended to services which comprise two thirds of our economy. This is NOT primarily a "free market" issue. This is a political issue, and politics will determine how it plays out.

posted by: camille roy on 08.25.04 at 04:53 PM [permalink]



Did anyone bother to listen to spencer? According to him, the "jobs created by outsourcing" are just likely already existing employees who get a new overseas boss!!! That would mean that that whole part of the analysis is junk!

Also Camille is right. Income inequality has always been a problem of political stability. You can push it pretty far before you get Bolshevism and Bastille day, but a disenfranchised angry middle class is exactly the right prescription you need to set it off.

posted by: oldman on 08.25.04 at 04:53 PM [permalink]



That is a brilliant analysis. Perhaps it can explain why unemployment in California generally is higher than in the country as a whole?
Perhaps it follows the arguement that the American economy is booming while there are 1.4 million fewer jobs available to 2.3 more Americans since Bush assumed office.

posted by: Mike on 08.25.04 at 04:53 PM [permalink]



gw writes: "Companies who do this could be made to pay for training for the laid off people to help them get into those higher-paid higher-skill jobs we keep reading about."

That'd be interesting.

Tell them to pay for 2 to 4 years of school (potentially at a decent college, not just community colleges) for each person laid off.

(Community colleges do a decent job, but they tend not to be cutting edge. If you want the careers of the future, you'll have to go to the places where that research is being done. )

posted by: Jon H on 08.25.04 at 04:53 PM [permalink]



I dispute that offshoring creates the same or greater number of quality middle and upper middle class jobs that it loses in the U.S. And I'll add that of the IT jobs created, no records are kept to determine if the workers are American or lower-wage imported non-immigrant visa (NIV) workers on H-1b or L-1s.

See Chris Nolan's eWeek article today for more comments and "Lost Your Job Yet?" in Computerworld at:
http://www.computerworld.com/careertopics/careers/story/0,10801,92150,00.html?SKC=careers-92150

posted by: John Pardon on 08.25.04 at 04:53 PM [permalink]



I'm also a software engineer myself (in embedded systems) with 11 years of experience.

No, I don't think there's any difference between programmers and software engineers other than the degree. And other than the class that taught C programming, none of the coursework required to get that degree has been useful on the job. The book, Practice of Programming by Kernighan and Pike was a lot more useful than concepts like recursion.

And in my opinion (not widely shared among my peers for obvious reason), what we do is ridiculously easy compared to other fields of engineering. It doesn't require any advanced math or science; you just need to know how to tell a computer to jump through hoops. And that's a learnable skill which I think is more comparable to writing HTML than engineering. So I think our salaries would be coming down even without offshoring.

Which is partly why I maintain a relatively low standard of living and save whatever I can.

posted by: fling93 on 08.25.04 at 04:53 PM [permalink]



Amen. Folks, there's nothing special or sacred about programming jobs. The skills required are for the most part rudimentary and are in any case fleeting.

The high salaries seen four years ago were a freak result of a one-time capital spending binge in this country that will not be repeated. It would be absurd to base any economic policies on protection of such a tiny, unsustainable, rapidly-obsolescent part of the workforce.


posted by: lex on 08.25.04 at 04:53 PM [permalink]



" I remember one case where we liked the person, but he was being paid contractor rates and was asking for a full-time salary of $150k. Believe me, that was not the prevailing wage for the kind of job we were hiring for."

In one of Yourdon's books, he states that an analysis of application developers determined that some developers are as much as ten times more productive than the average developer. $150K would be a bargain for one of them. It's a matter of supply and demand, and you just demanded mediocrity.

posted by: germ on 08.25.04 at 04:53 PM [permalink]



Fling93: "It doesn't require any advanced math or science; you just need to know how to tell a computer to jump through hoops. "

That's crap! You can't tell a computer to do what you don't know! I got the blank stare from an excellent programmer who couldn't see the equality between the theoretical formula for standard deviation and the machine formula, even though I derived it before her eyes. She proceeded to code to the theroetical formula, not understanding the concept of propagation of errors and the ultimate impact it had on her implementation.
Programmers code to a design; software engineers create the design.

posted by: germ on 08.25.04 at 04:53 PM [permalink]



"Programmers code to a design; software engineers create the design."

But that's not a job description. It's more like a state of mind or standard of practice.

It certainly isn't determined by one's job title. There are "programmers" who create designs; there are "software engineers" who are coding to a design created by someone else.

posted by: Jon H on 08.25.04 at 04:53 PM [permalink]



You can't tell a computer to do what you don't know!

There are always degrees of ability and knowledge. Indeed, there are plenty of programmers who are better designers than some software engineers.

I got the blank stare from an excellent programmer who couldn't see the equality between the theoretical formula for standard deviation and the machine formula, even though I derived it before her eyes.

I know plenty of degreed software engineers who would have given you an equally blank stare, or who think a hash table is the answer to everything, or who don't read documentation and take semaphores inside of a vxWorks signal handler (argh! and that was done by someone who was pretty senior).

And I've met very few software engineers who were good at writing a design document -- or that good at writing at all.

posted by: fling93 on 08.25.04 at 04:53 PM [permalink]



Good commenting.
Will add: good jobs = high rated sector of national defense and about as important as military defense in my view.

Free trade,invisible hand, etc.? Archaic outdated univ model (and tenured/socialists --while giving away all ideas to & condoning immigration invasions) have hurt the country too. Don't think those theories should rule the well-being of the country and its people. No, not Soviet style planning either.

posted by: Alex on 08.25.04 at 04:53 PM [permalink]



So, there's hope for those who lost jobs in 2001 and haven't been able to find work in their field since, and have not had an income to blow on worker-paid re-training or skills advancement?

I mean, despite the fact that a lot of them would not have been unemployed in the first place if they hadn't been unceremoniously replaced by managers who made bad business decisions and had to make up the lost costs (how else?) by laying off scads of workers?

Why is it, exactly, that replacing expensive American labor with cheap overseas labor can become such a fad that any company today that was once over 50 employees has done it... but replacing expensive overpriced commercial software and high-end hardware with free software? Is running Linux really more frightening than having your own software development and testing done by people in another country that you've never met?

Anyway. If someone can point me to the jobs listings board for these new on-shored tech jobs, I know plenty of heavily underutilized, very smart people who would love to apply.

Doesn't any free-market libertarian get concerned that we have lots of smart people in this country that aren't being put to work (because, say, the jobs arent there, or because the jobs that are there don't make up for the expenses incurred by taking them)? Isn't that, like, an organizational resource waste of a national magnitude?

Poverty's up, insurance is down, median household wages are down...

I know that having workers earn less (compared to inflation) is supposed to be good for the economy, but... if good economy doesn't mean people have more money, then what exactly *does* it mean?

posted by: kt on 08.25.04 at 04:53 PM [permalink]



More competition means a lot of things. It increases the supply of labor which, all things being equal, would push down the price paid for that labor. This has to be the dominant dynamic here. And when the price of that labor falls that means that the price of the service provided, in my example banking, also would fall.

Of course, the fall in services never runs in step or ahead of the fall in disposable incomes. Incomes fall first, then services feel a crunch because less people can buy those services. But this has to go on for a prolonged period before the service provider will lower services to make them affordable -- and will likely fund the lower price by providing lower wages to their own employees (or cutting labor rolls), who were not directly affected by the lower incomes of the customers until now. As a result, those employees themselves now have less money to buy other services.... and so on. Meanwhile, common expenses that are not driven by the domestic market -- such as energy -- may rise.

Fundamentally the problem right now is that: wages are falling, but expenses are not, as you suggest they must.

The personal view says... I want to know that my family are not going to be evicted.

This is important, because no one will hire you if you don't have an address or phone number. Try it sometime. (Actually, one of the liberal-funded programs that the Right perennially seeks to eliminate will find you something.)

All the concern about outsourcing programming jobs rings a bit hollow when one considers that so many programmers (or software engineers) employed today are foreigners, hundreds of thousands of them on temporary work visas

Not really. There has been an anti-H1B furor in tech circles for a long time. Companies post bogus job descriptions with unattainable qualifications, then claim that no Americans can be found to fill the position. They then acquire an H1B worker, who does not meet the qualifications either, but works cheaper than a red-blooded American. Unfortunately, more people have now lost their jobs to offshoring than to H1Bs, so H1B has lost focus.

Here's the thing. We've been complaining for almost 5 years now that companies have been underemploying, short-selling, avoiding, replacing, and overworking American workers. Replacing us with labor we can't possibly compete with only made the situation much worse. It's only now that it's come to a huge income-hitting head that mainstream moderates and conservatives and other laypeople have noticed the complaining.

The statement in the report that "programmer" jobs were lost while "software engineer" jobs were created reflects indeed more a trend towards fancier titles.

This, too, has been a trend for some time. It has been excruciatingly common for workers to be "promoted" with fancier titles -- and perhaps more responsibilities and/or hours -- but nothing else (no raise, bonus, perks, etc.) (It's called 'increased productivity', and it's good for the nation.)

Note, too, that the article makes absolutely no comparison between the wages of the programming jobs vs. the software engineering jobs. As Jon H has attested, there really is no tangible difference between the two terms in common use -- which means there's absolutely no reason to believe that these new 400,000 software engineers make more than the old 300,000 programmers -- not even 3/4s more, which would be the necessary factor to make this fact believably "good for the economy" i.e. household income.

But no, Dan (as well as the Bush administration) has shown repeatedly that it is only number of jobs that make a good economy, and not actual money being earned.

there would still be downward pressure on US programmers' wages because programming skill is a depreciating asset.

That's both true and not true. Knowing how to program is not a generally depreciable asset (any more than knowing how to add and subtract is). Knowing a certain programming language is depreciable, however.

The problem, of course, is that inept technical management thinks that the latter is equal to the former, and therefore requires specific languages instead of holistically assessing raw development ability. You can be an excellent programmer -- but if you don't know Enterprise J++.NET, well, "we'll let you know."

(Of course, we have a lot more programming languages than we probably need, and companies keep making more of them in their war for software development licensing share. And again, along the same lines I said before about companies not using free software, companies are strangely awfully resistant to develop in free languages and instead shell out for overpriced IDEs, compilers, and token methodologies)

But it's just not true that there was a large pool of unemployed American programmers standing by in 2000 waiting to be paid "market rate"

Have you tried that today? Try it right now. There are now large pools of unemployed American workers standing to be paid *less* than market rate. Of course, that's still not cheaper than overseas labor.

Don't free markets imply free movement of labor forces across borders?

Hah! We can't even get medicine from Canada -- because (gasp) it's too cheap! So you can forget about moving labor.

Can you imagine if we freed up economic restrictions and movement restrictions on affected U.S. families so they could move to another country where expenses are lower? We'd lose all that skills superiority that we're so proud of (but so unwilling to employ)... That would be *bad*. Besides, people who flee the country in a time of war (irrelevant of it also being a time of poor economy) are likely to be terrorists.

Free trade doesn't mean *workers* actually get to move around freely. Just goods and services and other sorts of things that the upper class can make money from.

posted by: kt on 08.25.04 at 04:53 PM [permalink]



kt writes "That's both true and not true. Knowing how to program is not a generally depreciable asset (any more than knowing how to add and subtract is)."

Agreed. It's basically methodical problem-solving.

"Knowing a certain programming language is depreciable, however."

What's really depreciable is knowledge of particular development tools, libraries, and class frameworks. Those come and go pretty quickly. Especially nowadays. It's hard to keep up.

Just in the case of Java UI libraries, there's been AWT, IFC, JFC/Swing, and now SWT.

Unfortunately, learning the ins and outs of these things is much more labor-intensive than learning a language. They can involve scores of classes, along with idioms and design patterns.

(Which is why businesses hiring based on *languages* is pretty absurd.)

"There are now large pools of unemployed American workers standing to be paid *less* than market rate. Of course, that's still not cheaper than overseas labor."

Alas, if you try to ask for less money, they'll think there's something wrong with you. So lowballing yourself isn't really an option.

And damn, it really makes it hard when a recruiter asks what you're looking to make.


Also, a last note on "Software Engineers" - degree programs in Software Engineering aren't very common. Most "Software Engineers" probably have CS degrees, just like many programmers.

posted by: Jon H on 08.25.04 at 04:53 PM [permalink]



"I look forward to the California state legislature's efforts to impose a tariff on services from Arizona."

Dr. Drezner knows perfectly well that this is unconstitutional. He also knows that an American programmer could in essence follow his/her job from Aptos to Arizona. He probably doesn't know that an American couldn't, even if he/she wished to put up with the radically lower standard living vis-a-vis public goods, immigrate to India. I have read that India only allows people of Indian descent to immigrate, (and a google search reveals no way of emmigrating to India, but does offer a lot of schemes to get Indians into the US/Canada/Australia as well as Indians bitching about illegals from Bangladesh) . Just one of the many, many flies in the ridiculous case of free traders. I look forward to when Dr. Drezners research grants and courses are outsourced to a hard-working Indian instructor with some grasp of reality.

As always in these matters, it is interesting to note the correlation between ethnic background and policy position.

posted by: stari_momak on 08.25.04 at 04:53 PM [permalink]



Can anyone cite a historical precedent for free trade? Is there a parallel between manufacturing in the 1980s and technology today? If there is no precedent in historical economics for a country utilizing a free trade policy, then it appears that this is an experiment with the future of America in the balance.

posted by: Dan on 08.25.04 at 04:53 PM [permalink]






Post a Comment:

Name:


Email Address:


URL:




Comments:


Remember your info?