Back in the early 1990s, I was working for a medium-sized software firm who had decided to try transferring its development to India.

At the time, the "world-wide web" was just getting started, and one of the first cartoons available online was Dilbert. As webmaster, I contrived a small script which would make today's Dilbert available from our website, with a click through to the main site.

Just as the Indian transfer was warming up, Adams was doing his "Elbonian" series of cartoons, where Dilbert's company is transferring its software development to a mythical apparently-Eastern-European country, Elbonia, which has no technological infrastructure.

Upper management was not amused at the correlation.

I have no issues with outsourcing per se. Economically, I'm convinced it when it's chosen correctly, it ultimately works out for the best for all involved -- except perhaps the poor schmo who must be retrained in something less outsourceable, and even that, only in the short run. Including if that's me, which it appeared to be.

But what I saw in the Indian outsourcing effort was mind-bogglingly stupid. No serious attention was paid to infrastructure costs, connectivity fees, actual labor costs, nor quality. It was as if the thinking was simply: "Wow, I'm screwing someone here in the US, therefore I must be making a bigger profit!" It was sort of an inverse parody of certain extreme leftist dogmas, where the belief that if someone is making a profit, someone else must be hurt.

Of course, they weren't increasing profits, and there was no reason to think they would be from the outset.

The VPs I'd discuss it with would say, glowingly: "There are more Ph.D.s per capita in Bangalore (or some other city) than anywhere in the world!" I tried to gently point out that sounded a little too good be true, and suggest perhaps some of those degrees came in cereal boxes -- that we were looking at the Lake Woebegon syndrome writ large. (No insult to India, but no country has that high a percentage of it's population as PhD-type material, as we'd think of it here in the US -- much less with decades of software development experience.)

At the outset, the Indian consulting company (Tata) promised us 3 qualified, degreed programmers for the equivalent of one US programmer -- by the end, the ratio was something pathetic like 4 to 3. And let's not get into hardware, facility, and networking costs. I could see this was probably costing more than it was saving long before we got into it, but upper management would not be dissuaded.

So we moved a software product to India, and discovered, after many months, what most companies discovered: (1) While there are indeed some excellent Indian programmers, our management was just too stupid to make sure our project was staffed with them. (2) Indians tend to be agreeable, and tell you things are just fine, even things aren't going well. (3) Costs were higher than American development. (4) US employees got sick of staying up till two in the morning in order to have conference calls. (5) When we were done, we'd paid millions for a snarl of code which didn't actually do anything.

So, since then, I've made much of my income from failed Indian outsourcing software projects, bringing them back to the US. Often companies just seem to assume that overseas software will be a lot more robust than turns out to be (A friend suggests they absorb that thinking from the trade mags.) In the worst cases, they lose millions (and more importantly, years) and get very little back in return.