The whole sending-development-projects-to-cheap-offshore-programmers meme is getting rather thick these days.
There are two very important economic truths to remember here. First, the high-wage "developed world" does not have a monopoly on brilliant people; there are some stunningly smart people living where the cost of living is much less than it is in high rent districts like Sillycon Valley, NYC and London. (Ziggy's Corrollary to Joy's Law: most of the smart people live somewhere else, too.)
Second, the kinds of jobs that are subject to wage arbitrage are commodity jobs -- not the kind of rewarding, high wage jobs that really need someone talented, but the kind of jobs that need a warm body with a pulse to wrestle with the computer until something works. These jobs used to be here, and they were crappy jobs here. When I worked in NYC, these were the kinds of jobs where immigrants would toil away for less-than-market-rates doing busy work because, well, no one capable would do them. (The only prospect for advancement here was for the "manager" who was responsible for ever-larger teams of "programmers" and billed the client for ever-larger sums.)
Here's some food for thought
What's getting outsourced? The IT equivalent of coal mining jobs. My undergraduate degree is in metallurgical engineering and I've spent some time in mines. Its dirty, but high paying work that doesn't require much formal education. People love it. But its also subject to lots of ups and downs and over the years had steadily declined.And this:My prediction is that while hundreds of thousands of IT jobs will go off-shore in the next decade, we'll gain more than we lose as we move up the hierarchy. We do a poor job of meeting demands at the top of the hierarchy and there's plenty of work to do. When you think about the real problems that IT should be solving, its amazing how little attention we pay to them.
-- Phil Windley on IT's Coal Mining Jobs
The project ââ¬Åuses natural-language understandingââ¬Â which, last time I checked, more or less amounts to being able to pass the Turing test, which a bunch of the smartest people in the world at MIT and Stanford and so on have notably failed to do, and it seems just a little unlikely that this bright shining goal can be offshored to wherever the cheap programmers are this year.If I had a googlable copy of Future Shock handy, I'd point out the passage where Alvin and Heidi Toffler talked about the bogeyman of mechanization in the 1950s and 1960s. They both used to work on an aircraft assembly line, one of the first kinds of jobs that got mechanized. The Tofflers reminded us that no one should shed a tear over these jobs, because these are the kinds of jobs that nobody actively wants in the first place.-- Tim Bray on Offshore BS
It's one thing to tighten bolts for a living because the the nation is at war. It's a completely different thing to expect a long career for the rest of your life as a "bolt tightener" because that was the skill you learned during WWII.
Right now, the IT sector is in the process of weeding out the "bolt tighteners". The challenge for us now, as Phil points out, is to find and solve the problems that need to be solved.
Re:Mobile capital, technological change, and worke
ziggy on 2003-12-24T00:53:34
Yes, as a large-scale trend, loss of US jobs offshore is a very devistating issue that no one is really trying to fix in the US. I find it utterly depressing that Wal-Mart is now the largest employer in the US.As for whether this process hurts or helps US workers, I think it's a toss-up (sure, you lost your manufacturing job, but look how cheap all the stuff is at Wal-Mart!). Sometimes, there are very clearly disastrous social effects of radical capital shifts: think US Rust Belt (Gary, Indiana), or the equivalent area in northern England.Yet I see the drivers for outsourcing IT work to other companies. In some small companies I've worked for, there just isn't enough money to hire capable IT staff. IT is a service, a business expense, and doesn't contribute to the bottom line. Furthermore, if a company like this can afford to hire someone junior, they certainly cannot afford to keep their skills up to date. And let's not talk about what happens when something really tricky comes down the pike, like a nasty Outlook worm.
So, on the one hand, consolidating jobs and outsourcing IT services to a contractor is a net benefit: access to highly skilled, capable practitioners, pay for what you need (even if it's not a FTE). On the other hand, there's a net loss of (entry-level) jobs, but a net benefit to business and skilled workers who want a fair wage and a challenging job.
Then there's the trend to send the outsourced jobs offshore to East Elbonia. As I stated above (and below), I don't see this trend holding up to sustained scrutiny. On the whole, offshoring of IT jobs isn't like the maquiladoras, because the kinds of low-paying jobs that are most easily sent offshore are the ones that are made obsolete by new technology. And creating new technology is what we do best here in the high-wage world.
I think you are wrong on two counts here:
First the kinds of jobs that can validly be outsourced are jobs that I want to do. I know because I telecommute from another continent, albeit one with a cost of living that's similar to the US. If I can do it from afar, so can someone equally smart from a cheaper country. And I like my job and I don't think it's a crappy one. It's just that it can be effectively done by anyone with a telephone and a good internet connection
And even if you were right, if only crappy jobs could be validly outsourced, haven't you worked in corporations before? Once the trend has started, they will outsource everything, even it doesn't make sense. Sure it will be non-optimal, but by the time enough PHBs realize it, you, I and many others might have been out of a job for long enough that we wouldn't be in this industry any more. A policy doesn't have to make sense to be applied. And I don't buy the usual "but in this case you shouldn't work for this company anyway": I have worked, and enjoyed working, for companies that made really questionable decisions in some areas, but which still provided me with a very good job.
The archetypal job that I do not see being outsourced is salesman, because you need the face to face interaction. Would you rather be a salesman, selling software developed in a cheaper country, than the guy writing that software?
Re:not quite...
ziggy on 2003-12-23T23:53:29
I think you're confusing outsourcing with the everything-is-moving-offshore hysteria.Sure, there's going to be lots of outsourcing all over the map. It's not just the blue collar cleaning jobs, but also programming type jobs as well. It's been happening for years, and will continue. The difference is that most of the outsourced jobs in IT are going to IT contractors and consulting firms that concentrate in something, like hosting, development, network administration, whatever.
Those jobs aren't going to low wage countries to arbitrage a wage differential. This type of outsourcing acknowleges that every company doesn't need to hire and retain top-flight software developers and sysadmins to run a network of 25 PCs and keep the email flowing. Some of that outsourced work will move to lower wage areas, like from California to North Carolina, US to Canada, or England to Ireland. But for the most part, the value is in concentrating specialist expertise, not finding a replacable cog that does the same job for 10% of a full time employee.
Yes, that's true, but for the most part, those aren't the jobs that are moving offshore en masse, like shoe production did for Nike, or textile production did for New York City's garment industry. Those are the jobs where a company finds the one or two people capable of performing the job, and structuring a telecommuting program for them. That's totally different from offshoring, which tends to focus on replacing a staff of 25 entry level programmers in the US for 100 entry level programmers in East Elbonia who work for 1/10th of the pay.First the kinds of jobs that can validly be outsourced are jobs that I want to do. I know because I telecommute from another continent, albeit one with a cost of living that's similar to the US. If I can do it from afar, so can someone equally smart from a cheaper country. And I like my job and I don't think it's a crappy one. It's just that it can be effectively done by anyone with a telephone and a good internet connection.I think you're taking this reductio ad absurdum a bit far. Yes, companies are going to replace IT workers with consultants and contractors. There are lots of jobs it doesn't make sense to send offshore, and lots of important IT functions fall into that basket. But you're presuming that all corporations are mindless lemmings that follow this week's trends.And even if you were right, if only crappy jobs could be validly outsourced, haven't you worked in corporations before? Once the trend has started, they will outsource everything, even it doesn't make sense. Sure it will be non-optimal, but by the time enough PHBs realize it, you, I and many others might have been out of a job for long enough that we wouldn't be in this industry any more. A policy doesn't have to make sense to be applied. And I don't buy the usual "but in this case you shouldn't work for this company anyway": I have worked, and enjoyed working, for companies that made really questionable decisions in some areas, but which still provided me with a very good job.Not every project is going to be sent to East Elbonia; there are still lots of projects around that benefit from having everyone working under the same roof (or in the same city, or in the same timezone, or on the same continent, or
...). And as Tim Bray points out, this whole wave of offshore hype isn't going to magically crack problems that are currently beyond the grasp of staff in high wage countries. Finally, I think you're unfairly discounting lots of companies that actually value their employees, and place a big emphasis on hiring the best people they possibly can. This approach is inconsistent with dumbing down all work until it can be sent out to the lowest bidder, possibly thousands of miles away. Maybe this offshoring craze is just a form of darwinian selection to cull out the companies that do manage by following this week's breakthrough management strategy.
:-) Re:not quite...
mir on 2003-12-24T00:32:39
I think you're confusing outsourcing with the everything-is-moving-offshore hysteria.Yes I was confusing the 2. My main point though is that no matter how effective outsourcing or offshoring really is, if it becomes trendy to do it, then companies will _have_ to do it, otherwise they stock will suffer. And anyway the managers will go to seminars, or listen to HPCs (Highly Paid Consultants) that will tell them that that's what they should be doing, and eventually they will.
I believe the 2 are very similar BTW, both stems for a desire to get more for less: use underpaid subcontractors, whether local or remote instead of full time employees.
Re:examples?
ziggy on 2003-12-27T20:53:08
A good portion of government and defense related IT jobs cannot be moved offshore. Some of that work is done by offshore subcontractors, but certainly not all of it. And there is a limit to how much can be sent offshore due to political, legal, security, or privacy issues. With added regulations on patient privacy, medical IT may soon fall into the same category in some countries.Can you give me a few examples of IT jobs that cannot be moved to another country, and why?Exclude the ones where physical and cultural proximity is important -- requirements, sales, UI design, customer hand-holding.
There are economic drivers in play here, too. While the German government may send lots of Euros to the US for software licenses, there is also a huge incentive to keep software development in-country. Many of the same issues are in play in Canada. Some of these factors are "soft cultural issues", but certainly not all of it.
Ultimately, the reason why IT jobs cannot be moved entirely offshore are tied to physical proximity, responsiveness, quick turnaround, or all three. Many projects need fast turnaround times, or need onsite resources.
I've worked on projects over the years where being in the same city was just barely acceptable, and the project would have cost less, been delivered faster or delivered with fewer bugs with an on-site developer. I've also worked on projects where the developer and customer were in the same timezone, ~100mi apart, and the distance and limited face-to-face communication and impeded progress. Similarly, on projects where team members are separated by 3 or 5 timezones, coordination becomes adds a noticable drag to the project. Separating the team by 8+ timezones adds a lot of drag to the project.
I've also worked on projects where development on a "date driven project" was agonizingly slow with the entire project team under the same roof. Sending part of the work to an office 2 miles away was just unacceptable, whether that was in the same company or a contractor. And I've seen projects where proprietary knowledge was involved, and sending any work outside was not an option; sometimes work was sent to another office (with cheaper labor), other times it was kept in the same office (with more customer involvement).
This is not to say that all development should always be done in house; it's more an observation that the futher the participants are separated, the more drag on the project. And many times, some or all of this drag is just unacceptable.
Now, I don't claim that my experiences paint the totality of IT work. I do claim that there are enough disincentives against offshoring for a great many projects, based on the limited amount of situations that I have experienced personally. This does not invalidate the need or desire to send work to Ireland, Australia, India or Iowa, but it does tend to invalidate the proposition that all IT work will inevitably flow to the provider with the lowest labor cost.
Re:examples?
brev on 2003-12-28T00:26:44
Hang on a minute. I thought the gist of your post was that offshorable IT == lower-skill IT. So I asked for examples of IT jobs that were so high skill they could not be exported.
Now your objections are more about the whole outsourcing concept, which applies within as well as outside the USA. (Hawaii is quite a few timezones away.)
Did I misunderstand your original post? Otherwise I'm still waiting for examples.Re:examples?
ziggy on 2003-12-28T05:49:07
Not quite. I started out by saying, «the high-wage "developed world" does not have a monopoly on brilliant people...». Martin Fowler recently wrote about ThoughtWorks, and their experiences with development centers in Bangalore and Melbourne. For example, he concludes (in part):Hang on a minute. I thought the gist of your post was that offshorable IT == lower-skill IT.Fowler's experiences are guardedly positive. ThoughtWorks has a high set of standards for hiring programmers, whether they are in the US, India or Australia. They are hiring abroad because it increases the chances of them hiring the best people in the world. However, this is not the standard, simple minded «let's send all the work to India because programmers are cheaper» mantra, which is facile.As I write this, offshore development is very fashionable, but it's still too early to really understand its true strengths and pitfalls. Certainly anyone doing because they think they'll get cost savings similar to the rate differences is seriously deluding themselves. [...]One conclusion is clear, anyone who thinks that onshore developers will triumph because they are more skilled is very wrong. We've found that we can hire just as talented developers in India as we can in North America and Europe.
I can see where you thought I was equating low-skill IT with offshorable IT. That was not my intent. Rather, my primary point was, «the kinds of jobs that are subject to wage arbitrage are commodity jobs [...] the kind of jobs that need a warm body with a pulse to wrestle with the computer until something works».
Mostly, I'm arguing against the fashionable meme that all IT jobs inevitably will migrate to Upper Elbonia because their labor costs are 1/200th of what there are "back home" (wherever that may be). One reason why those jobs won't (or shouldn't) go offshore is because outsourcing (or outsourcing to team that's geographically too far away) doesn't make sense for lots of projects.
One very important reason why those jobs won't (or shouldn't) go offshore is because cheap labor -- even in large quantities -- is not enough to solve hard problems.
[continued next post]
Re:examples?
ziggy on 2003-12-28T05:59:10
Actually, I was arguing the inverse -- it's not that high-skill jobs will not be sent offshore, but rather the jobs that will be sent offshore in pursuit of lower labor costs are low skill IT jobs.So I asked for examples of IT jobs that were so high skill they could not be exported.For example, accounting systems are a very well-understood domain for IT. Yet each large company's accounting system has unique wrinkles. At the very large end of the spectrum, it's not a problem amenable to a generalized solution. Yet it is not something that requires a Ph. D. in computer science, or tight integration with the customer. It's a low skill kind of project that used to be the bane of entry level programmers, interns and co-ops. Today, it's also the kind of project that's amenable to sending to a team in Upper Elbonia where you can expect a satisfactory result.
But you asked for the high skill jobs that cannot be exported. That's a fair question, and I did avoid the issue previously. So here are some examples that should help paint a picture of what's "high skill", and not amenable to sending offshore:
The common threads I see is that many of these problems are not IT-bound; that is, domain knowledge (finance, genomics, supercomputing, high energy physics, legal regulations) is more important than IT in some of these problems. Many of these projects are not amenable to outsourcing or long-distance outsourcing, and therefore not amenable to offshoring. Many of these projects need highly skilled labor (both IT and non-IT), and cannot be made cheaper by sending the code out to the lowest bidder (supercomputing applications, 3D rendering, managing big data, etc.).
- Defense related projects
- nuclear weapons research
- military software (missile guidance, tracking systems, autonomous control)
- New and innovative applications (pushing at the edges of engineering and CS):
- 3D Rendering (e.g. Pixar vs. DreamWorks for Hollywood movies)
- Finance (trading, arbitrage, derivatives)
- Video games
- Bioinformatics
- Telephony
- Government and Regulatory Applications
- Taxation
- Regulatory Compliance (e.g. Sarbanes Oxley in the US)
- Medical Records and Privacy
- New Drug Applications
- Public Records
This list is certainly not complete, but I hope it's enough to paint a sketch. I've seen the "commodity jobs" that require gobs of "warm bodies" to complete. One of them was a re-engineering project outsourced to a big consulting firm. According to their Gantt chart, it was 16 months into the project, and time to start hiring programmers. In the interview, they actually said to me, «We've done all the thinking. All you have to do is sit down and just write the code. When can you start. Do you have a resume?» And that's just one kind of "make-work" job -- there are plenty more.
Re:examples?
brev on 2003-12-28T23:05:34
I think I see what you were saying now. So the limits to outsourcing have to do with the separability of IT from domain knowledge. Since the domain knowledge resides in the USA, the more interesting jobs will reside in the USA.
That may be true for the short term, but I'm not sure about the long term.
- one day soon if not already, important customers/markets will not be American.
- even for American customers, domain knowledge isn't all that tied to the USA. American companies consult around the world too. Are these problems insuperable for them?
- outsourcers in places like India are not going to stand still. Every system they build, some of the domain knowledge transfers overseas. Today, Indian programmers do 'bolt-tightening'. Tomorrow, look for them to want to move up the food chain. They'll do whatever it takes to get there, and all they have to do is be slightly lower in cost.
You're speaking to a guy who's half-Indian and has a lot of relatives in software and engineering. (I'm one of the least engineerish of the whole family, actually.) Over the holidays this has been a hot topic at my parents' house. Some of my relatives were doing outsourcing before any of you ever heard of it. One cousin has been offered a very tempting position back home, much better than anythiing available here -- but his family is a bit too entrenched in the USA now to consider it. So you can see that the next generation will have a much easier time doing a 180 upon graduation from the prestigious American schools, and taking all that cutting edge talent back to India.
Our general perception is that what India can do is very much underestimated. even today.
Anyway, this doesn't necessarily invalidate your point -- maybe some things will not be easily outsourced. But at that point, I expect Indian companies to emerge that can swallow the entire project whole.
Re:examples?
ziggy on 2003-12-29T15:37:52
That's a good chunk of it. Some work will remain here (or in Canada, or Switzerland, orSo the limits to outsourcing have to do with the separability of IT from domain knowledge. Since the domain knowledge resides in the USA, the more interesting jobs will reside in the USA....) because there's a critical mass of learning and people available to crack a problem. There's no reason why India or China can't become the leading center of computational astrophysics, but NASA Goddard will probably remain one of the leading institutions for some time to come, just because it takes so long to gather all of the expertise together -- from the theoretical physics to building and running beowulf clusters, to writing massively parallel applications. India and China will probably be peers with the likes of NASA Goddard, CERN, and other centers of theoretical physics at some point in the future. Maybe they'll surpass the existing labs. But India's gain won't be America's loss. Even if India becomes the focus of this research, that doesn't mean that everyone at Goddard is out of a job because of cheaper wages elsewhere. I suspect that situation will be more common than bleeding-edge jobs moving en masse to India.
I never said that Indian programmers are limited to low-tech jobs like 'bolt-tightening'. I said that the best kinds of jobs to send offshore in pursuit of low wages are the ones that have meager domain requirements. I'm also very aware that Indian programmers are also doing some cutting edge work, like those at ThoughtWorks. Those kinds of jobs aren't sent abroad in pursuit of lower wages, but in pursuit of talented workers.Today, Indian programmers do 'bolt-tightening'.Remember, I'm arguing against the meme that all work will flow to India based on lower wages alone.
Certainly.Tomorrow, look for them to want to move up the food chain.Not quite the same, but...
htoug on 2003-12-28T15:53:23
A good portion of government and defense related IT jobs cannot be moved offshore.Nearly all our Goverment and defense related IT here in Denmark is done by a US company - OK most og the workers are danish, but the owners and the management is american.
It bothers me quite a bit, but it doesn't seem to bother the goverment, who happily sold our CPR (Central Person register, cnf. the US Census and the registry of births and deaths rolled into one - it contains your current address social status and so on and on and on...), the Taxation programmes, the medical records of everyone and nearly all defense applications (the ones that ensured that the danish soldiers in Iraq are properly equipped with grassmovers, snow-shovels and salt to remove the ice on the wintry pavements in Basra!) to CSC -a US coporation.
OK, it's not outsourcing the IT jobs, it outsourcing the management, but I think that it's even worse.
Re:Not quite the same, but...
ziggy on 2003-12-28T19:03:09
Right, this is neither outsourcing nor offshoring IT. There are similar situations all over Europe, where foreign companies (usually American) are employing locals to do work for the local market. The jobs are staying in-country, not moving to Elbonia because IT costs are cheaper there.OK, it's not outsourcing the IT jobs, it outsourcing the management, but I think that it's even worse.The same thing is happening in the US: instead of direct employment, lots of jobs run by contracting organizations that hire local staff. Sometimes, the contractor (e.g. IBM Federal Services) hires the customer's former employees to do work for the customer.
We see the same results, except on a smaller scale. Profits from a contract in Pennsylvania pay for staff in Pennsylvania, but mostly go back to the home office in California, New York, Washington State, or elsewhere.
