Arnold Kling  

Finite Labor Supply in India

PRINT
What Explains Inequality?... Academic Self-Selection...

Anti-outsourcing alarmists say that India has an "infinite supply of cheap labor." If that were true, then we would not be reading stories like this.


According to a Nasscom Hewitt Total Rewards Survey 2003, the entry level salary (junior software engineer) is up an average 24% to Rs 267,000 from 217,000 in 2002.

Thanks to a reader for the story tip. Outsourcing continues to be one of the top economic stories of the year. See Daniel Drezner's recent post. However, my guess is that over the next 18 months, we will either see a stronger U.S. labor market, a slowdown in outsourcing due to higher wages in India, or both. Most of all, we will see the end of election season. Being a free trader in an election year is like being a turkey a month before Thanksgiving.

For Discussion. When Council of Economic Advisers Chairman Greg Mankiw said that outsourcing is not different from other forms of trade, there were calls for his resignation. If President Bush were forced to hire a chief economic adviser who opposes outsourcing, who might he pick?



COMMENTS (17 to date)
Tom Dyess writes:

Sure, they have a finite number of potential employess. Unfortunately that finite number exceeds the finite number of U.S. employees by over 400%

Bernard Yomtov writes:

It strikes me as poetic justice that a Bush economic adviser gets harshly criticized all around for telling the truth.

triticale writes:

If President Bush were forced to hire a chief economic adviser who opposes outsourcing, who might he pick?

H. Ross Perot?

Eric Krieg writes:

Lou Dobbs. Seriously, have you watched his program lately? You would think that CNN is thinking of replacing him with Lougeev Dobbsia or something.

He supposedly has a degree in economics.

Cap'n Arbyte writes:

Paul Craig Roberts?

(ducking!) :)

Mcwop writes:

On the other hand Marginal Revolution has a post about "insourcing". This adds more balance to the debate. POST LINK

Boonton writes:

Probably the administration wouldn't replace the economist with an anti-outsourcing one. They would just replace him with one who didn't tell the truth (that outsourcing is a good thing) and just avoided the subject instead.

John Thacker writes:

Tom-- the sheer number of workers is not the point, either. The US has ten times the population of Canada-- that doesn't mean that Canadians don't have jobs, or that companies don't have workers in both places.

The point is that Indian wages will rise to meet the marginal value of labor, and the outsourcing trend will stop.

david foster writes:

Bear in mind, though, that other countries will be coming on stream as suppliers of software development...China, Phillipines, Russia...and, at the same time, those increasing salaries will encourage more people in India to study computer science.

One factor that is rarely mentioned is that SW development isn't a very capital-intensive activity. It's not like setting up an auto plant complete with expensive stamping presses et al. This tends to make it much easier for work to follow the lowest costs (a similar phenomenon has been seen in the apparel industry (more so than textiles, which is more capital-intensive)

India's exports of IT services & software to the US in 2003 = $7 billion.

Total size of IT services & software market in the US (2003) = $500 billion+

Need I say more?

Don Boudreaux writes:

I know the point is trite, but until someone can explain in a compelling way how the creation and enforcement of monopoly power is a general means of increasing an economy's productivity, protectionism will never be anything other than a self-serving excuse for security against wealth-creating market forces and free consumer choice.

Lawrance George Lux writes:

Free Traders are always quick to state Indian Wages will go up, but decline to comment that American Wages will have to come down in response to outsourcing. The quoted statistic is an anomalty. The retained American software engineers(fewer in number) are paid more for their speciality and experience. It does not state the total loss of software jobs in the industry.

I vote for Arnold himself, so he can deal with the real world. lgl

John Thacker writes:

Free Traders are always quick to state Indian Wages will go up, but decline to comment that American Wages will have to come down in response to outsourcing.

Again, the reason not is because: 1) Prices and wages are relative, and 2) To the degree that outsourcing raises productivity everywhere (which is does), it raises wages everywhere.

Over the long term, so long as American productivity doesn't decrease, American real wages will not either.

ed writes:

Hmmm.

1. Anti-outsourcing alarmists say that India has an "infinite supply of cheap labor."

For all practical purposes they do. Higher education in India is heavily subsidized by the government so IT workers there don't have the enormous college debt that IT workers here in America have to deal with. Additionally pool of available workers is growing every year simply because the average age of that pool is relatively low.

Another major point is that the vast majority of jobs going to India are of the junior-programmer variety. The reason why this is dangerous is that there is now developing a gap between relatively novice level jobs, such as junior-programmer, and senior level jobs. Without that intervening layer of jobs it will be very difficult for young American programmers to advance themselves to the next level of their careers. At some point American companies may be forced to hire Indian IT workers simply because the level of expertise and experience could not be found domestically.

2. Total size of IT services & software market in the US (2003) = $500 billion+

The vast majority of that comes from sales of software or of licenses and so those will always count as domestic sales regardless of where the software was actually created.

Should it surprise you that much of WindowsXP was written by Indian programmers in India and Redmond? Does it surprise you that all domestic American sales were attributed to the American division of Microsoft?

So if an American company generates sales from a product produced by non-Americans, according to you, we should take that a sign of healthy employment of Americans? I could use a rude phrase here but I'll simply send it telepathically instead.

I think you can read my mind on this one. :/

3. Again, the reason not is because: 1) Prices and wages are relative, and 2) To the degree that outsourcing raises productivity everywhere (which is does), it raises wages everywhere.

This is patently absurd.

While this homily sounds good it is little more than nonsense. Outsourcing raises wages in the destintion for outsourcing, not in the origin. At the point of origin it in fact depresses wages by forcing application of the lowest common denominator. And if you want to refute me, use logic and facts please. No nonsense. Any more homilies like the one above and I'll treat you very roughly.

4. Over the long term, so long as American productivity doesn't decrease, American real wages will not either.

If you're talking about the federal productivity figures then I've got to point out that such numbers DO NOT include the effects of outsourcing. I.e. if company ABCD used to produce 1,000 bits with 300 employee but now produces, by outsourcing, 2,000 bits with 50 employees it will show a massive increase in productivity.

Which doesn't seem to be warranted eh?


In summary I personally think that much of the pro-outsourcing debate is complete bullsh*t. There are precious few facts, a great deal of wishful thinking and an enormous quantity of referencing the odd dead economist. Frankly I haven't seen all that many ecnonomists still living get things right so I hope you'll understand my disdain for the reverential treatment of dead ones.

For those of you who are pro-outsourcing here's a couple things for you.

1. 4% annual GDP growth.
2. 9+ months of recovery now past.
3. 16,000 jobs (only!) created in Dec 2003.

Frankly I think the theories are funny and the quips marginally less so. But the final result is the one that is important. IMHO 2004 will see the first recovery with a net loss in jobs. If you disagree, and I certainly hope you do, then expound and explain in as explicit a fashion as possible.

Be creative, be diligent and include an absolutely painful amount of detail. Vague "wait till October then you'll really see the job creation" posts will be ignored.

Joy writes:

ok - let's get this fact straight...there is no longer a connection between GDP growth and job creation. Why, you ask? Because it's a GLOBAL economy, duh! GDP can increase in a country, but jobs won't necessarily be created in THAT country.

I actually discussed this with the folks at the BEA (Bureau of Economic Analysis). They are fully aware of the disconnect.

Too bad no one in the Bush administration has a clue.

Sam writes:

The comment that software development isn't very capital intensive shows that the individual has never actually been involved in software development.

First of all, in order to develop software, one needs an education. That costs money.

Second of all, one needs a machine on which to test the developed software.

Third, if one expects to deliver the kind of highly integrated, expansive software applications widely in use today, one must purchase any number of the following:

- Third party libraries.
- An operating system.
- A database server system.
- A web server system.
- An application server.
- An e-mail server.

Even if you go with "open source" systems, you must still invest in the intellectual capital necessary to work with those systems. Just those things alone can run into the millions upon millions of dollars.

That doesn't even mention the office space and other "tools" necessary to do the job.

Sure, it might not be a metal stamping machine, but Microsoft Visual Studio ain't cheap either.

Comments for this entry have been closed
Return to top