Thanks for commenting on my book Hive Mind. Two responses:

1. Scott Alexander has rethought his IQ-by-decile critique, though he hasn’t withdrawn it. The top 10% of countries ranked by IQ earn 12 to 20 times more than the bottom 10% (12 vs. 20 is driven by how you treat Botswana, an important outlier). So even with Scott’s decile approach, IQ predicts much more prosperity for nations than individuals, even in a simple bivariate test. Scott obliquely discusses his rethinking here: http://slatestarscratchpad.tumblr.com/post/134882180601/another-dumb-stats-question-to-test-your-patience#notes and he sums it up indirectly here http://slatestarcodex.com/2015/12/10/list-of-passages-i-highlighted-in-my-copy-of-hive-mind/ though there’s more on Twitter.
2. My academic paper with Schneider on the magnitude question finds that 1 IQ point predicts 0.5% to 1% higher wages for individuals, but 6% or 7% higher productivity for nations. So the R-squared approach, the decile approach, and the regression coefficient magnitude approach tell similar stories: IQ matters more for nations than for individuals. http://mason.gmu.edu/~gjonesb/Immigrant%20IQ
3. On the rounds-of-experimental cooperation question: One of the prisoner’s dilemma games is an average of 4 rounds, while the one I coauthored lasted 10 rounds. And the McCabe trust game is just two rounds! So if we’re thinking about political party leaders cutting deals, I suspect the number of rounds isn’t too short. In relatively short-run games, higher group IQ tends to predict more cooperation. [The first game I mention is here: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2505361]

Thanks very much for taking the time to read my book.

-G