In some ways, this post dovetails with William Easterly's paper, "Was the poverty of Africa determined in 1000 BC?":
http://aidwatchers.com/2010/07/was-the-poverty-of-africa-det... ?
Notice this, from the post:
"1500 AD technology is a particularly powerful predictor of per capita income today. 78 percent of the difference in income today between sub-Saharan Africa and Western Europe is explained by technology differences that already existed in 1500 AD – even BEFORE the slave trade and colonialism.
[...]
Why do technological differences persist for so long? The ability to invent new technologies is much greater when you have more advanced technology already. James Watt had acquired a lot of tech experience in the mining industry which he used to invent the steam engine. Other people with the ability to make steel could then slap his steam engine on a vehicle running along steel rails and give us railroads."
Part of his (sobering | frightening | surprise) conclusion is that early advantages quickly become cumulative advantages that can persist for very long periods of time. The roots of Western Europe's success in the 18th and 19th Centuries might be in 1500 or even 1000. You have to look very far back in time to realize that. Likewise, the roots of Silicon Valley's advantage, according to the post above, might go back at least a century.
That's not the kind of advantage you can uproot with some tax credits, a five-year forecast, and a fistful of cash. You might need to play a game that works on the scale of decades or longer -- far longer than any politician or city booster thinks.
These kinds of ideas, whether about poverty in Africa or tech in Silicon Valley, shouldn't be taken as overly deterministic, but they should be useful to those who think about quick fixes for very wide, very deep problems.