Surely this productivity to cost ratio will change over time as more and more IT work is done in India. Look at the relationship in Mexican manufacturing, for example. Significant productivity gains over time with little to no rise in real wages. The benefits of production in the Global South are precisely that a higher rate of exploitation is possible.

But no worry, of course. There is no need for "low skilled" labor in the US after all . . . (pay no attention to that Wal-Mart behind the curtain!)

The anecdotal evidence presented in the article does not invalidate the fact that most programming jobs do not require tremendous skill and will certainly go offshore.

The point is that most IT engagements require a blend of onshore and offshore resources. The ratio for many projects is around 1:2 onshore:offshore, although the actual value, ie, pricing to the customer, may well be weighted toward the onshore resources. The point is that, while most of the VALUE may remain generated by very process-savvy senior professionals working in the US, most of the HEADCOUNT will go overseas.

In other words, there is increasing demand from corporations for the very high-priced, deep industry expertise of senior consultants who have a thorough grasp of arcane and complex processes like supply chain management. However, a project usually requires only a few such senior consultant types, whereas most of the actual coding can easily be done offshore.

Sum: Good news for industry vets and McKinsey-type consultants. Bad news for garden-variety 30-ish programmers. Good news for niche consulting startups. Bad news for huge staffs of mid-level IT specialists.

Real wages in Mexico took a huge hit from the devaluation that took place at the end of Zedillo's term, so any analysis that includes that time period will of course show falling real wages.

But if the increase in trade w/ Mexico brought on by NAFTA did not lead to an increase in labor costs then why are many Maquilas leaving Mexico and transplanting operations to the Pacific Rim?

The truth is that, since the devaluation, real wages in Mexico have risen.

The problem in Mexico is that productivity has not risen along w/ wages, perhaps due to inefficient investment in infrastructure.

General Glut,

I wouldn't disagree with your conclusions about who reaps the rewards of productivity increases. There has been a huge windfall for the owners of assets during the last 10-15 years, with little increases accruing to those who do not own stocks, houses or companies. Even within high-tech companies, the gains accrue overwhelmingly to a few top insiders. Two months after I joined my last software company, the COO, who'd been hired a scant 20 months before me, cashed in several hundred million $ worth of options. Within a year, he left the company.

The real debate should not be about how to prevent jobs going overseas--they will continue to do so--but about how to more equitably allocate ownership so that workers can participate in the financial gains generated by productivity improvements and growth.

At the same time, we should make it as easy as possible for worker bees desperately trying to hang onto back-office and IT jobs that in all fairness should go to India or be replaced with smarter processes and better technology to start their own small consulting companies. Nothing more closely aligns one's product/output with what the economy demands than the need to offer it in the marketplace.

I'm not an expert on tax or healthcare policy, but surely there must be plenty of ways we can appropriately encourage potential startup owners so as to encourage people to add real value to the economy rather than continue to fight rearguard, losing battles against Indian 24 year-olds.

I hear you've been named Outsourcing-R-Us. I would say that fits pretty well.

What? You mean pre-trained Indian contract workers are too vertical to be of any use to a forward-thinking company who needs dynamic workers?

And that their long daily schedules lowers the chances of them being able to gain new skills while they are doing your work?

And that the overhead of dealing with an agency that is 8,000-11,000 miles away from you, and on a day schedule 8-12 hours off from yours, makes it really difficult to dynamically alter the workers on the other end?

Coulda probably told ya that. Myra and the rest of us thought the exact same things. But we're just geeks, not MBAs, so we wouldn't be smart enough to predict that.

Let's see if it holds.

I really have to wonder about Daniel's grinningly sarcastic introduction to this post. After all, up till now, he has been defending the disingenuous pro-offshoring mantra that "getting the cheapest labor is good for our economy". But this story does not defend that. In fact, it says that getting the cheapest labor is NOT good for our economy, because the cheapest labor cannot meet our productivity and innovation needs. It is a blip on the objective anti-offshoring radar that says that limitless, knee-jerk offshoring DOES NOT WORK -- which is not what DWD's pro-offshoring line has been.

I guess I can't complain if cookie-cutter programming jobs, previously held by New Media/New Economy terps whose expertise consisted of a "Visual J++ in 24 Hours" book under their arms, get moved offshore, and said twerps get the sales-manager jobs they really should have had in the first place.

But there still remains the crime of the offshoring world that it was so completely uninvolved with the operation of its own business that it thought so much of its programming work was that menial that it could be written down, wrapped up nice and neat, and sent across the world into a black box. No one actually working in those sorts of jobs could ever possibly think that.

And I'm sure every programmer and other dynamic, advanced IT worker that got offshored said that to their bosses, too. I know we did.

Hey Dan,

Where are the high-skill employees going to come from, to handle those complex tasks in the US, if all the entry-level jobs go abroad?

Huh, genius boy?

Keith -- au contraire, I most certainly do not advocate outsourcing über alles. Outsourcing makes sense only when the marginal revenue product (output per dollar of labor) overseas exceeds what it does in the U.S.

I agree that some of the outsourcing we've witnessed has been done because of management fads. In fact, I said in Foreign Affairs that:

It is also worth remembering that many predictions [of huge amounts of outsourcing] come from management consultants who are eager to push the latest business fad. Many of these consulting firms are themselves reaping commissions from outsourcing contracts. Much of the perceived boom in outsourcing stems from companies' eagerness to latch onto the latest management trends; like Dell and Lehman, many will partially reverse course once the hidden costs of offshore outsourcing become apparent.

I'm claiming that done properly, offshore outsourcing is good for the American economy and does not destroy jobs. Critics of ffshore outsourcing critics are simultaneously claiming that it's an inefficient move and that it's an irreversible trend.

Outsourcing can be both a fad and a trend, depending on who does it and why. It can also be a product not of some interested consultants advice but of management's desire to make short-term numbers look better by cutting costs, something that may not help a company in the long run but will bolster management's case for increased executive compensation in the short run.

There are in fact lots of companies that have outsourced functions -- customer service being the obvious example -- that management does not value or understand, because a reputation for outstanding customer service helps a company only if it is sustained for many years longer than the typical executive expects to be around.

Where are the high-skill employees going to come from, to handle those complex tasks in the US, if all the entry-level jobs go abroad?

Tsk tsk. You sell Danno and his fellow travellers short! They've created the perpetual motion machine of wage depression: we can't do the work in the U.S. because we have no skilled people - we have no choice but to outsource. We can't produce experienced people because we have no low-level work for them to cut their teeth on. You get a death spiral. One can imagine the hand-to-forehead wailing of poor, beseiged CIOs, forced, weeping, to send work to India for want of a few good American programmers.

The ones in this country that will really make out are the trade and business schools that push overpriced training courses for Microsoft and Cisco certifications. They peddle these courses at ten or fifteen thousand per to the young and foolish, promising them "$60 to $70K!" when they're done, only to have them find out the paper without the experience is worth bubkes.

"the paper without the experience is worth bubkes."

Actually, experience is worth bubkes if it isn't recent. If an IT person hasn't been able to find work for a year or two, they probably won't get hired at all, even if they have 10 years of experience before being unemployed.

Dan wrote: "Critics of ffshore outsourcing critics are simultaneously claiming that it's an inefficient move and that it's an irreversible trend."

These aren't mutually exclusive.

If offshoring is inefficient, it'll take a while before businesses figure that out. For a while, they'll try different approaches, or different service providers, thinking it's a problem with their processes and management, not an inherent problem in offshoring. If offshoring doesn't seem to work, their instinctive response will be to offshore harder.

In the meantime, they're letting people go in the US, or not hiring new employees. So US tech people either aren't finding jobs out of college, or aren't finding positions after being laid off.

Companies that are hiring IT workers aren't interested in people who have no experience, or who haven't worked in a while. IT people have a very short shelf-life, in the eyes of the HR people.

So people are dropping out of the field in school, and laid-off people are being pushed out of the field because they can't get hired, due to length of unemployment. (Or, for that matter, due to their age).

So fewer people will be working in IT in the US, and fewer people will be entering the profession. Low-level work will be sent overseas, and only people with high-level skills will be hired. But US people with high-level skills will become more and more rare, because US techies will have few opportunities to learn high-level skills through experience.

So US companies, having acted in ways that discourage people from entering the IT industry, and having set things up that prevent US people from gaining the high-level skills and experience they require, will have no choice but to increase offshoring of jobs.

This because all the high-skill, high-experience people will be appearing in India and elsewhere, rather than in the US.

Simply recruiting US employees with MS and PhD in CS is unlikely to suffice. Academic work doesn't necessarily translate directly into real-world development skill. It'd be like having sex with someone who's got a PhD in human sexuality, but is a virgin and has never seen a naked human up close.

Jon H.,

The pattern you describe appears in other lines of work. I'm a trial court research attorney specializing in law & motion - the volumes of pre-trial motions in civil cases.

Law firms which specialize in insurance defense (chiefly personal injury) have been so nickle and dimed by the insurers that they ceased training programs for associate attorneys, with the result that fewer and fewer insurance defense counsel even know how to win, or limit/shorten, civil cases before trial. They're losing motions through incompetence which they should have won. Trial courts are being flooded with trials of civil cases which should never have survived to trial.

Another example appears in other threads here - the Democrats really don't have people who know beans about national security. President Clinton's first Defense Secretary, Les Aspin, was the first to mention the problem. Fewer and fewer bright young Democrats went into the field after the Vietnam War because almost all preferred domestic issues.

By 1993 there was a horrible dearth of suitable Democrats for Aspin to appoint to medium and high level DOD positions.

Things later got so bad that Clinton had to appoint a Republican senator (Cohen) as defense secretary when Aspin got sick and died.

Richard Clarke's ego and lack of professionalism will result in a total wipeout of Democrats in national security positions. At this point the only place Democrats will be able to get any national security experience will be on Congressional committee staffs.

Only they're in a minority in both houses and Senator Jay Rockefeller's blind partisanship with his intelligence committee staff has caused the Bush administration to increasingly stiff Congressional committees on classified matters.

One of the two major purposes in the assignment of covert operations to military forces is to avoid Congressional oversight of the CIA. Congressional Democrats are letting that happen because they just don't care about such oversight duties.

So there are lots of vicious circles aka positive feedback events.

“I'm claiming that done properly, offshore outsourcing is good for the American economy and does not destroy jobs.”

A spoiled yuppie like Jon H. could care less about the logical arguments for job outsourcing. He has been brought up to believe that the world owes him something. The idea that he may need to retrain for other work is met with great hostility. This is why the Democrat Party is so attractive to immature people like himself. It thrives on attracting “victims” of one sort or another.

General question:

Is it to soon to look at this through productivity numbers? I ask because I've seen lots of things get outsourced in the past... mostly to domestic outsourcing for IT support... and even when the employees don't change, merely where the place the checks come from changes, productivity as a rule goes down.

Such productivity losses tend to get worse in direct proportion to the degree of process and personel change.

Is it therefore unrealistic to not expect large productivity losses... enough to offset any cost savings, for the first several years of the new org being in place?

If you look at this chart from the Economist stored here:

http://www.nathannewman.org/log/archives/001689.shtml#001689

And you read the words of the Economist - hardly considered an anti-globalist publication! Nathan coommenting writes:

One culprit is outsourcing-- not just the raw fact that jobs overseas are cutting employment in the US, but a statistical problem. Since internal shipments within firms, especially services done at a distance, don't always show up in trade figures, sales in the US reflecting that foreign labor may incorrectly imply higher production actually in the US:

Then he quotes from the Economist
For example, when American firms outsource call-centre and information-technology-support jobs to India and other Asian countries, the result should be higher imports of services, yet official statistics do not show such an increase. America's recorded imports of software services from India are also much smaller than India's reported exports of such services to America.

It goes back to that sticky productivity problem. Productivity gains are artificially inflated by not properly counting the inputs from foreign offshoring.

So speaketh the Economist.

"It comes down to being where the customers are." But "where the customers are" is subject to change. For example, as manufacturing activity moves offshore, then "the customer" (ie, the decision-maker) for ERP systems, robotics systems, etc...moves as well.

An issue often overlooked in the offshore outsourcing debate especially of programmers is that of H-1B visa holders. During the tech bubble, programming jobs in the US were often filled by people from India, China and other countries coming to the US on H-1B visas. Since companies couldn't hire enough people even so, the cap on the number of H-1B visas issued per year was increased. By the end of 2001, 890,000 people were in the US on H-1B visas. Even in 2002, when the job market in the US was not doing well at all, 200,000 applications for H-1B visas were approved (many of them renewals or transfers, but at least 79,000 new issues).

It is not obvious how to interpret these numbers. There are apparently not enough Americans (or permanent residents) available to fill all these jobs, although H-1B opponents are claiming that employers use H-1B visas to keep salaries low (or at least lower than they would otherwise be).

Many job cuts in high-tech companies over the last few years did affect H-1B visa holders in large numbers. Since they don't count as unemployed (AFAIK), this also skews the number of jobs lost. I wouldn't be surprised if a few hundred thousand jobs previously held by H-1B holders have been destroyed since 2001. One can argue over whether this matters since supposedly there hadn't been any Americans/residents to take those jobs anyway.

The fact that I keep mentioning permanent residents has several reasons. A lot of H-1B holders eventually apply for and are granted a Green Card (and thus they become permanent residents). Moreover, in the high-tech companies I have worked in, most of the programming jobs not taken by H-1B visa holders were taken by permanent residents, not Americans.

I also have first-hand experience of the complete and utter failure of an offshore outsourcing project, just as described in the New York Times article. The ratios of cost and productivity were almost exactly as in the article - we were paying for a group of 10 people about as much as we would have paid for about 3 people here. These 10 people were producing, in terms of quantity alone, about as much as 1 or 2 people would have produced here.

But it gets worse. First, managing this group was a rather high-maintenance task and at times consumed more than half of the time of two project managers here. (And they had their own project managers, too - one here and one in India - so it wasn't as if this project was under-managed.)

And the quality of their work was horrible. This was not entirely obvious at first and only became apparent over time as we tried to maintain what they had produced. Most of what they had produced in India ended up being thrown away and/or rewritten.

Why was this so bad? I don't know. Certainly not because they were Indians - almost all the programmers working so much more efficiently here on the American end were themselves Indians on H-1B visas.

Even as the failure became more and more apparent, our upper management didn't get it. Our CTO confided in me that we were getting such a great price on that team in India and that he was thinking of extending them for another six months even though their price would go up a bit. I pointed out their productivity ratio to him which made him fade a bit. In the end, their contract was not extended, but it was mainly because we had to cut costs. Nobody even considered letting more people here go to keep the offshore team - they might as well have shut down any serious development effort.

It is not inconceivable that most of the good Indian programmers who are reasonably proficient in English (an important requirement) actually decided to come to the US on H-1B visas during the tech boom. Many managed to convert their visas into Green Cards or at least Green Card applications (which allow them to stay until the application is decided upon). It is also possible that by coming to the US, these Indians quickly adapted to essential differences in work habits and local customs that are important for a successul project.

Another aspect to consider: outsourced programming projects frequently fail, whether the contracting company resides within the US or outside of the US. Possible reasons include that the contractors don't know and/or don't care enough about a project which they are going to hand off to someone else within a few months.

I'm not writing this to make any particular political point. I just think all these issues have not received enough attention in the offshore outsourcing debate. It's true as Dan and others have written here that outsourcing has become a management fad and is done without much regard to its likelihood of success. I'm not sure high-tech outsourcing takes jobs away from Americans since there aren't enough Americans available to do these jobs to begin with. However, one should revisit the question of H-1B visas and ask what role they play in the outsourcing debate. Should H-1B visas be reduced or completely done away with? Or are they, quite on the contrary, the right way to do "outsourcing", i.e. by getting highly educated people from all over the World to come to America, improve their (and their families') standard of living dramatically and spend some of their money here as well?

It goes back to that sticky productivity problem. Productivity gains are artificially inflated by not properly counting the inputs from foreign offshoring.

I'm not sure I'd limit that trend to offshoring. Consider my earlier example of domestic IT support outsourcing. Before stuff being outsourced, many internal functions wouldn't ever get recorded/reported. A question I've long had; how much of the so-called increased productivity of the 90's wassimply more reporting on what was already there? How much of the so-called dot.com bust was people finally figuring that one out?

Outsourcing makes sense only when the marginal revenue product (output per dollar of labor) overseas exceeds what it does in the U.S.

That's a general argument. Apply it to the current reality of uber-low costs of labor -- ANY labor, skilled or otherwise! -- in other countries, and you have a result where the general argument covers quite an awful lot of the possibilities.

I agree that some of the outsourcing we've witnessed has been done because of management fads.

So what? Fad or not, it leads to a lower bottom line, which increases corporate profit, which en masse improves the economy, right?

I'm claiming that done properly, offshore outsourcing is good for the American economy and does not destroy jobs.

So you defend the practice, without examining whether its actually being done properly? I may not be looking hard enough, but I have yet to see you criticize a company for "bad" offshoring, or even mention "bad" offshoring before at all.

I don't think that the practice of moving jobs overseas (or the practice of promoting it) is a sure bet for a better economy. It's a lot more of a gamble than that. Sure, if America does a moderate amount of offshoring, and median household incomes (or especially income vs. expenses) don't drop like stones, the lower corporate expenses can encourage expansion in other domains, and even perhaps (one could hope, that is, if we lived in a sensible system) lower costs of goods, leading to greater domestic sales, leading to greater consumption, leading to a need for more geographically localized work, etc.

Maybe, but even in that rosy longview, there's idealistic assumptions; like on the behavior of executives with their profits, or that executives truly understand the sociologies behind their markets (or potential markets).

Regardless, I don't think that's happening.

On the other hand, if America does a knee-jerk amount of offshoring, putting or keeping hundreds of thousands of once-decently-paid (though typically overworked) skilled laborers out of work, letting median household income tumble, while common expenses soar, leading to a widespread increase in debt and increase of personal bankruptcies, leading people to cut household expenses (and therefore sales of goods and services, especially local ones), then (and please correct me if totally wrong here) domestically-sourced revenue will drop, leading to less investment in domestic services and marketing, leading to less need for localized labor -- not to mention probably more offshored jobs to make up for that lost revenue even further.

Critics of ffshore outsourcing critics are simultaneously claiming that it's an inefficient move and that it's an irreversible trend.

It's not irreversible -- labor tariffs would certainly reverse the trend. But it will certainly continue if it is left unabated. You said yourself in this very post that managers (and uh, executives) want to push the latest fad instead of the smart decision. (Certainly they are not pursuing utilitarian solutions!) This would support a belief that not only are managers making bad staffing decisions (i.e. based on popularity rather than critical analysis), but that they tend to do so as a rule.

Which is why to this day most skilled laborers so strongly dislike most managers, and why Scott Adams can be rich in both up and down economies.

GW said: These 10 people were producing, in terms of quantity alone, about as much as 1 or 2 people would have produced here. ...[M]anaging this group was a rather high-maintenance task and at times consumed more than half of the time of two project managers here. (And they had [two of] their own project managers, too... so it wasn't as if this project was under-managed.)

Frederick Brooks would not be the least bit surprised. In fact, he'd probably become aghast upon hearing you say just the above.

However, hiring and staffing managers, especially in IT, are so insanely convinced of the direct proportionality of the "man-hour" notion, that some will go so far as to fire anyone who mentions Brooks' book.

Undoubtedly the dogged skepticism over Brooks' theories of interemployee communication overhead -- or the inability to perceive or conceive of the concept -- plays a part in the drive by companies to offshore, at least those who did it in order to have a greater number of workers.

Apparently despite the fact that Brook's book, or at least the point of it, is incredibly well-known in the circles of skilled workers (especially IT), hardly anyone hired to manage engineering projects has ever heard of it, or accepts it.

Maybe offshoring will turn around. It might not happen until under-competent management is eliminated. In any case, I doubt it will happen soon enough.

Have you looked at PhD computer science student profiles at a school like Carnegie Mellon or Yale recently? About half of them seem to have completed their BS and MS from another country (Eastern Europe, India, China) and the rest of them seem to be from second-tier universities from the US. So the first group has the option of moving back home any time there is a slump in the job market. And no one from a "good" school wants to go into computer science any more.

I don't know what that means for research jobs in computer science (jobs that a PhD is presumably supposed to pay you for), but as someone above said, when the entry-level positions go to people in other countries, you can't just produce a brilliant developer of neural networks out of nowhere. It's all the same work, some of it has slightly more mundane applications and some of it goes into research and development and I don't see how the second part can stay in this country if the first part goes abroad.

And I hear little to none in these articles about the consequences of the export of technology. I read a long, long article by Brad DeLong and all he talked about was available jobs and jobs created but no one ever talks about the impact of technology. What will this country be like when industrial research isn't cost-effective enough for companies to fund here anymore? Are we so secure in our knowledge that we are technologically superior to every country out there that we don't believe they can replace us in research and developement jobs?

Or, I dunno. Is it just that people who write articles on "outsourcing" are service sector political/economic commentators who bypassed a career in the sciences for one in the liberal arts or social sciences and, like most of this country, don't value science.