Outsourcing: tempest in a teapot

Outsourcing is still a topic that can turn your typical American programmer into a fan of import protections. It's perfectly understandable, as we're talking livelihoods, and all the economic theory in the world won't keep the electricity on or pay high LA mortgage rates (which is why I won't be buying a house anytime soon). Of course, I'm a member of an American species rarer these days than an ivory-billed woodpecker. I'm a firm believer in free trade. Don't interpret "belief" as implying lack of proof, however, as we have 100 years of economic history that makes the theory more sound even than Darwinian evolution (which isn't saying very much if you're from Kansas).

But I'm not here to debate the merits, or lack thereof, of outsourcing. Rather, I question how big a threat outsourcing really is to American software developers.

A recent ZDNet article noted that Indian software and services exports are on the rise. A month ago, however, we were told that outsourcing was falling from favor, as companies discovered that expected gains from outsourcing didn't match the reality.

That seems counter-intuitive if you are stuck thinking that software development is a zero-sum game. In a zero-sum, fixed quantity universe, gains in India MUST result in losses to Western nations. If your kid brother insists on sitting over the dividing line in the back seat of the car, there's less room for yourself. You're left with no recourse but to throw your balogna sandwich at him, causing Dad to pull the car over and threaten to spank the living daylights out of you.

Fortunately, the software economy isn't a Ford Pinto driven by Dad (think what an exploding gas tank might do to financial markets). As the price of steel goes down, companies use more steel. As the price of apples go down, companies use more apples in food products.

Outsourcing lowers the price of software development. It does this, however, only in areas conducive to outsourcing. Products that can have all development managed externally, such as maintenance on existing products and additions to existing products, benefit most from the cost benefits of outsourcing. New software development, or custom software for internal use by a company, often benefits least from outsourcing. In the former case, the lack of instantaneous communication is less of a problem. In the latter case, instantaneous communication, the ability to change rapidly in response to new requirements, and an understanding of local needs are critical to success, making it a poor candidate for outsourcing.

In other words, most software development tasks can't be outsourced. Even so, lower costs serve to expand the software market in a number of ways. First, lower costs in themselves result in higher demand, just as lower cost steel yields more demand for it. Second, moving maintenance and upgrade tasks offshore leaves domestic developers more time to concentrate on new software development tasks. This productivity enhancements results in more software getting developed.

That's why Indian software companies can continue to grow while American companies get burned by some outsourcing projects. Markets must figure out how to use a new resource input, in this case outsourcing. Some use it well, others don't. While companies figure it out, though, the market is in disequilibrium, as tasks which shouldn't be outsourced are outsourced, displacing workers unnecessarily.

Companies are nearing the end of their experimentation phase. That's why I don't fear outsourcing.