Wednesday, September 21, 2005

Forget about software, go run a business

The blogger Mini-Microsoft is probably not the only employee who’s riled up by Business Week’s report of a new pay scheme which will supposedly go into effect at Microsoft this Fall:

“If they meet incentive goals, the 120 or so vice-presidents will receive an eye-popping $1 million in salary a year, and general managers, the next level down, will get $350,000 to $550,000, according to a high-ranking source. But the rest of the staff is paid at market rates.”

This doesn’t surprise me (assuming it’s true) since a flat stock price and competitive recruiting by competitors encourages exceptional rewards.  It would be disconcerting, though, if the exaggerated rewards only go to senior managers, as the report implies, and not to top engineering talent as well. The dramatic performance disparities in software productivity are well known; see my entry on Engineering jobs, which ruminates on the fact that tech companies can never get enough of the very best people, while the rest of us struggle to get by.

But then again… I’ve been learning about the “Law” of Comparative Advantage, which indicates that players which produce multiple goods are better off if they trade with each other, with each specializing in the good which it produces most efficiently.  India and China probably have a comparative advantage in writing software vis a vis the US, while the US has a comparative advantage in entrepreneurship and finance.  

Applying Comparative Advantage to the software business – in a rather cavalier way, I admit – suggests that US companies will specialize in business management, while India and China will write the software.  This would mean American managers and Asian developers.

(Note: I’m not saying that India and China have an absolute advantage in software; the law of comparative advantage only requires that they’re relatively better at software than they are at finance.  They could be worse than the US at both, and it would still be to both sides’ advantage to specialize in the way I’ve indicated. In this example: If American tech managers are three times as good as Asian managers, and American software devs are only twice as good as Asian devs, then you’ll end up with American managers and Asian devs. However, invoking Comparative Advantage in the way I’ve done is a stretch, since its simple proofs focus on labor productivity in widgets/hour, which really doesn’t apply in software.)  

Good news for Microsoft VPs and GMs, but not for Redmond developers.