Wednesday, September 22, 2004
previous entry | main | next entry | TrackBack (1)
The GAO's Rorshach test on offshore outsourcing
Over a year ago, U.S. Representative Adam Smith (D., Wash.) asked the GAO (which used to be called the General Accounting Office, but has since been renamed the Government Accountability Office) to study "issues related to offshore IT services outsourcing." As the offshore outsourcing brouhaha heated up, more and more congressman dogpiled on top of this request, expanding the GAO's mandate beyond just the IT sector.
The first part of that report has been released today. It's essentially a literature review of available government data on the magnitude and impact of offshore outsourcing. There are two themes that come out from this: 1) the government data on this phenomenon is incomplete and imperfect; 2) what data exists suggests that offshore outsourcing is not quite the tsunami it's been made out to be.
This is from the Results in Brief (p. 3):
And this is from p. 15:
This is consistent with my own back-of-the envelope-calculations from earlier this year.
Now, what's interesting is the responses to this report. This is a snippet from the press release by two Seattle-based labor unions, SPEEA-IFPTE and WashTech:
A tip of the cap from everyone here at danieldrezner.com to U.S. Representative Adam Smith. Beyond the unbelievably cool-sounding name, Smith has acted like a responsible grown-up on the offshore outsourcing issue. His one op-ed on the subject didn't demagogue the issue, and offered an eminently sensible, constructive request -- expanding coverage of Trade Adjustment Assistance to include service sector workers. No hysterical claims that offshoring was destoying the American economy, or even his district. Just a sensible policy proposal and an appropriate request for more information. Also, in contrast to the aforementioned unions, it appeared he's actually read the GAO report.
A politician who seems reasonably well-informed and resists scapegoating a non-issue. Damn, that's refreshing.
Oh, and for those who just can't get enough of offshore outsourcing, the United Nations Conference on Trade and Development (UNCTAD) has just released its 2004 World Investment Report. If you download Chapter IV, there's a nice overview of the offshoring phenomenon.
UPDATE: Brier Dudley and Marilyn Geewax have dueling stories at the Seattle Times and Seattle Post-Intelligencer respectively. One data point that captures attention is the fact that "the number of business, technical and professional services, flowing into the United States, however is rising, from $21.2 billion in 1997 to $37.5 billion in 2002," as reported by Geewax (this is CNN's lead as well).
That's an increase of 76.9%, which sounds really bad. But it's only half of the picture. What about exports of business, technical and professional services?
Those precise figures weren't in the GAO report, so I e-mailed their staff to see if they knew -- and they promptly replied. As it turns out, during the same period, exports of these services rose from $44 billion in 1997 to $64.5 billion in 2002 (This is from the Bureau of Economic Analysis's Survey of Current Business, October 2003, p.65, Table E).
So in other words, between 1997 and 2002, when offshore outsourcing is supposedly taking off, the balance of trade in the services likely to be offshored went from a $22.8 billion surplus to a.... $27.0 billion surplus.
My heart be still.
FINAL UPDATE: In fairness, see this erudite comment below by an IT consultant. I certainly won't deny that offshoring can have a hard affect on indivudual workers -- I just don't think it warrants the hysteria that, say, this comment epitomizes.posted by Dan on 09.22.04 at 02:49 PM
And he's a Democrat. SCORE!
(By the way, I didn't even catch the coincidence with the names until you pointed it out. But it is cool.)posted by: Brian on 09.22.04 at 02:49 PM [permalink]
Get enough of offshore outsourcing?
Actually, I'm waiting for the inevitable thread drift to Iranian nucks.posted by: TexasToast on 09.22.04 at 02:49 PM [permalink]
I recognize that the macro trends in offshore outsourcing suggest that the actual impact to the economy & jobs is far less than the hysteria. And wouldn't argue it.
This does not change the fact that offshore outsourcing has a very real and substantial impact in certain vertical markets of our economy. I’ve served as a Managing Director or CIO for 3 different international IT consulting firms. In these roles I’ve encountered and have had to counter or incorporate this impact every day in almost every one of my sales and execution activities.
Over the past 4 years offshore outsourcing has been the major contributing factor to the erosion of our billable consulting rates in percents ranging from 20 – 30 % for project managers and top-notch domain experts to 45 – 60 % for more “commodity” IT services – Java / .NET programmers, database developers, web developers, support and administration etc. Yes, the competition of a repressed IT market in general has also contributed to rate reduction as firms compete for a lesser volume of available work. But I can state on experience that the primary impact is offshore outsourcing because I have seen a least three identifiable cycles in this time period where rates fell & the market stabilized for a time as firms using high percentages of offshore content won bids on man-hour price, not experience or ability, and everyone else had to adjust rates to be competitive.
During the last 4 years I have seen offshore competition on bids shift from only encountered on high-dollar, specialty service bids to a regular occurrence on even small, multi-tens to hundred thousand dollar projects. As example, database developer rates have dropped from 60 – 75 $ / hr to 25 – 35 $ / hr on some bids. The only way I can compete on most competitive bids is through alliances and subcontracting to offshore firms. And let me tell you, on any substantive work the added management and coordination effort required to control scope creep and schedule risk makes even what used to be the most straightforward of projects walk a fine line between profit and loss.
Sorry, Dan. I’ve read your postings and know your passion for debunking the “sky is falling” impact of offshore on the economy at large. And I agree. But I’m here to tell you that your macro view comes with a massive caveat that is masking quite a few alarm bells where component sectors (or at least mine) are being decimated. It is healthy and right to keep this dialogue open. But in my opinion there is more danger in minimizing a potential impact by taking too broad a view than there is my extrapolating a specific trend to an overall market. At the end of the day Chicken Little just goes home hoarse and no one else is really hurt in the long run.
WRONG, WRONG, WRONG AND MORE MYOPIA AGAIN.
Ruining the country your grandfathers and great grandfathers/grandmothers worked so hard to develop. This legacy of the legacy of the too soft and lazy--yuppies and offpsring guppies.
SMART PEOPLE DON'T NEED COLLEGE PROFS EITHER. SMART PEOPLE SHOULD BE ABLE TO READ.
ADDES TO THAT, NEED REQUIRED FOR YEAR FULL-TIME NATIONAL GUARD TYPE TRAINING FOR ALL CITIZENS( to to learn to defend this country's facilities and ports on this soil [for several more reasons]) All yuppy and guppy asses included.
This country has massive econ problems, massive trade deficits, massive lack of exports, massive immigration problems, massive funny moneys as investments, inflated housing pricing ( with even the home repair products foreign made), and even the sports teams stocked with foreign made goods, etc. etc.posted by: Alex on 09.22.04 at 02:49 PM [permalink]
Isn't it interesting that there's an intelligent, well thought out post by Jon followed immediately by screaming by Alex.
I think Jon's correct. The macro look at outsourcing is positive. The micro, down to the impact on the individual who loses his or her job, is much more devastating. Thus it was in the textile industry moving to the south, and thus it will be for today's outsourcing.
There needs to be a way to alleviate the micro pain while the macro good is accomplished.
I'm not sure I see the impact of outsourcing's "decimation" of certain sectors as being quite as localized as Jon does. In the long run the best candidates to graduate to the more creative, higher-paid IT functions are those who have some experience with more prosaic programming, development and support work. As outsourcing puts downward pressure on compensation for this work and the field becomes less attractive fewer Americans will seek to enter it.
This is not a good thing, and has both economic and political implications. I fear that the mostly positive implications for the global economy (of important work being done for less money) tends to blind some economists to the mostly negative implications for the American economy and especially the political impact of outsourcing here in the United States.
It should go without saying as well that those inclined to dismiss concerns about outsourcing (Dan Drezner, this means you) find it much more congenial to wage arguments against dopes who post in all caps and end their rants with "etc., etc." This is too easy to be worthwhile.posted by: Zathras on 09.22.04 at 02:49 PM [permalink]
"In the long run the best candidates to graduate to the more creative, higher-paid IT functions are those who have some experience with more prosaic programming, development and support work. As outsourcing puts downward pressure on compensation for this work and the field becomes less attractive fewer Americans will seek to enter it."
Good. Maybe they'll do something else besides programming. There's an awful lot of creative innovative work that needs to be done - we're still burning the stuff that John D. Rockefeller was cranking out, and driving glorified Model T's, for crying out loud!
Of course what brought all this brainpower into the computing industry in the first place was that people could hire them to work on cool things and then see how they floated in the marketplace without wasting endless amounts of time playing "Mother May I" first, and people profited by hiring smart people to come up with good stuff, not by getting their buddies in Washington to write a new regulation to lock out upstart competitors. We need to apply the regulatory model that the computing industry (pre-Microsoft-antitrust-lawsuit) enjoyed to every industry, so that our smart people will have somewhere else to go once programming gets so easy that everybody and his dog can do it, which is already visibly starting to happen.
And then we'll get massive technological advances in something other than computers for a change. A true win-win situation.posted by: Ken on 09.22.04 at 02:49 PM [permalink]
Well, Ken, I wish I had your confidence that the energy sector is a likely destination for those discouraged from getting into the IT sector because of outsourcing. Frankly, I was thinking of programming and similar IT functions as a place the really creative, productive types could start on their way to more advanced tasks in the same industry. There's no reason to assume that without that foundation the project managers and CIOs of the future will all be Americans just because they are Americans.posted by: Zathras on 09.22.04 at 02:49 PM [permalink]
"Well, Ken, I wish I had your confidence that the energy sector is a likely destination for those discouraged from getting into the IT sector because of outsourcing."
Well, they're going to have to go somewhere. If there's lots of other interesting fields that have the same regulatory environment that computing does, they'll likely end up in those fields advancing non-computing technology. Otherwise, we miss out on advancement we could have had and they get stuck with boring, less lucrative jobs.
"Frankly, I was thinking of programming and similar IT functions as a place the really creative, productive types could start on their way to more advanced tasks in the same industry."
Assuming there continue to be more advanced tasks in that industry over the long term. I don't think that's a very good assumption. Programming's getting to be less of an advanced art as programming problems get solved, reusable libraries accumulate, VB continues to enable more people to be productive in the industry who wouldn't have been able to in earlier years, and improved hardware make tight memory and CPU usage progressively less important. The industry doesn't need as many geniuses as it used to, and some of them can move on to do other useful things like find a viable substitute for oil (i.e., one that doesn't involve everyone driving around in pregnant rollerskates, crowding back together in inner cities, and living primarily on local production), finally consigning the groundcar to museums where they belong, or working on anti-aging drugs. Deregulation across-the-board will make this transition easier and quicker, to everyone's benefit.
Embedded stuff will still take lots of talent for a while, until even the embedded hardware has the kind of speed and storage that PC's do today. As it is, virtual machines with garbage collection are viable on cellphones.posted by: Ken on 09.22.04 at 02:49 PM [permalink]
What I never see in the discussions of offshore outsourcing is an application of the old economic principle (due to Herb Stein, I believe) that what cannot go on forever, won't.
Indian IT salaries are now about 20% of American salaries, for comparable jobs. But the savings for outsourcing to India is only about 30%, due to communications and management overhead. And that's assuming you do everything right -- you may get no savings at all. This overhead is not likely to go down anytime soon.
Indian IT salaries are also currently going up about 15% per year, which means they will double about every 5 years. The corresponding American salaries are stagnant at best. This means that in 5 years Indian salaries will move from 20% of American salaries to 40%, and the savings of outsourcing to India will drop from 30% to 10%.
Again this is because the overhead costs of outsourcing are not likely to drop much during the next 5 years.
You can see that this situation cannot possibly last beyond the next 6 to 8 years. Indian salaries will either have to level off, or American salaries will have to rise. Most likely both will happen.
But 6 to 8 years is a long time in IT, so it's unlikely that India will stay satisfied with only taking in America and Europe's laundry. By 2010 you are likely to see more and more Indian companies striving to compete head to head with American and European companies.
India has a large well educated middle class that is hungry to rise in the world. But even India is not a inexhaustible source of software talent. Almost half the population is impoverished and illiterate. And even among the educated middle class, only a small fraction have what it takes to become a good software developer.
I've worked with Americans, and I've worked with Indians in both the US and India. The Indians working in America are a very elite bunch. The Indians working in India have about the same percentage of losers, screwups, and mediocrities as you find among a comparable group of American developers.
I also have to say that I have worked as a software developer for almost 30 years now, and one constant I have observed is that every 5 to 10 years large numbers of software workers get swept out of the profession because they either weren't that good to begin with or they let their skills stagnate. Many jobs that software professionals were paid to do one year become automated out of existence 5 to 10 years later. This has always happened and will keep happening.
This would have happened to a lot of Americans working in IT in 2000 even if no one in India spoke English or learned to do IT development. And the same fate awaits a lot of people working in Indian IT right now.
If you want to work in software development, you can never take your job for granted. This was true long before anybody ever heard of outsourcing to India.posted by: S. Anderson on 09.22.04 at 02:49 PM [permalink]
1) There are many extremely bright 30-something year old software engineers who are experiencing difficulties in the current job market - either long out-of-work periods or substantial income reductions. These people, no manner how bright, are NOT going to just turn around and discover a "viable substitute for oil" or "anti-aging drugs" instead - not unless someone invents time-travel so that they can go back and change what they studied and did between the ages of 18-25. Technical skills are simply not that fungible.
2) It used to be common wisdom in the software industry that a company's most valuable assets walked out of the door each evening. Historically, software companies whose products remained stagnant have gotten killed in the marketplace within a few years (Ashton-Tate/dBase being a classic example). I know of instances where IT companies were acquired solely to acquire the software development talent (in many examples, all of the company's products were discontinued and all of the company's upper management, accounting, and marketing staff were sacked). America's leading corporations are now building the talent base and infrastructure for a dominant computer industry in India, while allowing the talent base here in the USA to decline. If this trend continues, then it is almost inevitable that over the next decade America will start gradually losing its dominance of the software industry to overseas competitors - competitors that have, in effect, been trained by America's brightest IT minds. It is hard to convincingly argue that this is likely to be good in the long-term for the USA.posted by: JB on 09.22.04 at 02:49 PM [permalink]
"These people, no manner how bright, are NOT going to just turn around and discover a "viable substitute for oil" or "anti-aging drugs" instead - not unless someone invents time-travel so that they can go back and change what they studied and did between the ages of 18-25. Technical skills are simply not that fungible."
But the new crop of bright people will tend to avoid programming and look for something else to do. The 30-40 year olds tend to be more experienced than the guys in India, and still have a niche to fill in many cases. And recall that it wasn't just 18-25 year olds that switched into programming while it was still hungry for smart people. Let other fields grow the same way, in an absence of professional licensing requirements, and they'll be looking for high-IQ people with whatever current training to let them climb the ladder again in a new home. Sure their income will go down for a while, but they'll be better off than if they start working at Circuit City.
Like I said, we do need political changes to make this transition happen more smoothly, but none of them involve cutting off trade.
"America's leading corporations are now building the talent base and infrastructure for a dominant computer industry in India, while allowing the talent base here in the USA to decline. If this trend continues, then it is almost inevitable that over the next decade America will start gradually losing its dominance of the software industry to overseas competitors - competitors that have, in effect, been trained by America's brightest IT minds. It is hard to convincingly argue that this is likely to be good in the long-term for the USA."
But it is good in the long-term for the USA for us to gradually lose our dominance of the software industry. There's thousands of years of work for our brightest minds to do - if they're all doing software until the end of time, all the other good things that we could have gotten will never happen, and we'll be stuck on this planet until the next Hammer falls. (After that, there won't be much programming work for anyone for quite a while, if we're all still here)posted by: Ken on 09.22.04 at 02:49 PM [permalink]
"But it is good in the long-term for the USA for us to gradually lose our dominance of the software industry. There's thousands of years of work for our brightest minds to do - if they're all doing software until the end of time, all the other good things that we could have gotten will never happen, and we'll be stuck on this planet until the next Hammer falls."
How is it good?
How will the destruction of the US computer industry create more funding for pure research in other fields in America? It's not like the companies doing basic science research in the US have a severe shortage of qualified applicants. Removing high-tech jobs in one sector does NOT create them in another sector.
Also, one of the great things about software development is that people with 3-4 years of college education and an IQ moderately above average can be productive. With all due respect, such a person is going to be virtually useless in the sort of advanced pure scientific research about which you keep waxing poetic.
But hey, they can probably get a job as an assistant manager at Taco Bell.
"Let other fields grow the same way, in an absence of professional licensing requirements, and they'll be looking for high-IQ people with whatever current training to let them climb the ladder again in a new home."
You really don't know anything about the sort of advanced research you keep posting about, do you? The sort of pure science fields to which you keep referring require far more formal education than computer programming. Switching to one of these fields mid-career is virtually impossible.
I've know several bright engineers that have tried to switch to a new discipline mid-career (because the job market in their field was poor). Some of them spent several years making NO income while paying for expensive college classes before they gave up. I don't know of any that have succeeded (most ended up either perpetually unemployed or working for some place like Taco Bell).
I personally started as a physics major. I switched to software development because the job market for physics majors wasn't great (especially if you didn't get a doctorate). There is no way I could go back now. I've already forgotten most of what I used to know about things like differential equations, so it would be like starting over needing 5-7 years of college education. I'd lose my house if I tried.
Dan, you may be interested in this guys argument. Posted in URL.posted by: DM on 09.22.04 at 02:49 PM [permalink]
Computer Science and Engineering students are not reading Professor Drezner's optimistic analysis about offshoring because enrollments continue to plunge.
The decline has hit just about every type of school. At UC Berkeley,
Nationwide, new enrollments are at 1996 levels — and few expect them to
"It's been precipitous," said John Guttag, head of MIT's electrical
Post a Comment: