One of the challenging paradoxes during the past several decades is that American teenagers have consistently performed below average on international tests in math and sciences, and not especially well on reading tests, yet the American economy is more productive than any other.
Arnold's got some illustrious support. Here's P.J O'Rourke:
Why do some places prosper and thrive while others just suck? It's not a matter of brains. No part of the earth (with the possible exception of Brentwood) is dumber than Beverly Hills, and the residents are wading in gravy. In Russia, meanwhile, where chess is a spectator sport, they're boiling stones for soup.
The belief that education is important for development, I reckon, is a form of category error.
Put it this way. You might argue that I'm richer than the average Englishman because of my greater human capital. But I'm richer than the average Somalian simply because I'm English.
What's more, it's prodigiously hard to show that education is a big cause of aggregate income growth. Problems include:
1. Reverse causality. Let's say a country were to anticipate faster GDP growth, say because it was opening up to trade. Its people might then be more willing to send children to school, in the hope that education would enable kids to reap the fruits of growth. Faster growth would then follow increased education, without education causing growth.
2. Rent seeking. Education might prepare people to rip off others, rather than engage in genuinely socially productive activity. In a corrupt society, a law degree might teach you to play the system - and give you valuable contacts. It'll be good for you, but it won't add to GDP growth.
3. Non-linearities. Imagine you were the only person with your language who could read and write. What would you read? Who would you write for? Your literacy wouldn't be much use. Maybe, then, education causes growth only after a critical mass (pdf) of people has been educated.
4. Multiple causality. A government that's educating its people well is likely to be doing other things well too - like having enough rule of law to collect taxes, caring about people enough to give them some freedom. All these things can promote growth. So how do we isolate the effect of education?
Here's a classic paper (pdf) which shows just how hard it is to find a big role for educational differences in explaining differences in national incomes. Here's contrary evidence that education does affect growth, within OECD countries.