Paul Graham's entire argument rests, as far as I can tell, on the assumption that the distribution of great programmers maps to the population distribution. Can

Paul Graham's entire argument rests, as far as I can tell, on the assumption that the distribution of great programmers maps to the population distribution. Can this be empirically established or refuted? Because doing so would bring this matter to a close pretty quickly, I think. After all, is anybody arguing about whether America needs to import middle eastern oil? Of course not, because the global distribution of oil has been empirically established, and the need to meet local deficits with imports meets little, if any, resistance. I feel that the intractability of this immigration debate, such as it is, is due primarily to the fact that the global distribution of top quality "digital talent", as Graham calls it, hasn't been established. Some people say most of the best is out there, others say the best is just a matter of training.