The Sh*t L*st

I came across the following remark in the commentary for a blog entry on outsourcing:

> I'm generally in favor of free trade, but I'm hedging my bets
> by going back into defense work and operating my software
> business on the side.
>
> BTW: The pain felt by the software professionals is acute
> because so many of them were paid well beyond their
> abilities and worth. I have been cleaning up lousy code
> produced by a group of such 'wonder-kids' for the last three
> years.

This guy sounds a little resentful. But it seems like the wonder-kids' lousy code was a windfall for him. For had they taken the time to delouse the code, they would have been taking from him, in advance, the work -- and reward -- that he is now earning. I guess the case could be made that had the lousiness of the wonder-kids' code been unearthed earlier, they might have been canned sooner and mop-up here brought in sooner, or paid a higher wage according to his merits.

But then what are the larger ramifications of lousy code as compared to clean code? By automating processes, or wrapping up projects, that reward workers lower down the socio-economic ladder, clean code acts as a tax against labor on the margin. Lousy code spreads the wealth. Otherwise, the code cleaned, the project completed ahead of schedule, the extra profit accruing from these developments floats right up the hierarchy to executives and shareholders. And a couple cushion-warmers in Sector 7 may now be expendable.

Of course, completely malfunctioning code threatens to disrupt production downstream and occasion greater displacements. For society, the optimal outcome economically may be lousy, minimally functioning code that requires regular maintenance. But then isn't this Microsoft's business model?

Have there been any studies or treatises on the economic benefits of inefficiency?