Income and genetic inheritance

Bryan Caplan notes:

This 2002 paper by Bowles and Gintis reports that identical twins' incomes have a correlation of .56, versus .36 for fraternal twins. Using standard formulae, this implies that genes explains 40% of the variance of income, family environment 16%, and non-shared environment 44%. A recent working paper by David Cesarini gets income correlations of .545 for identicals versus .266 for fraternals, implying that genes explain 56% of the variance, shared environment -1%, and non-shared environment 45%.

It's easy at this point to say, "Of course! Intelligence has a large effect on income and is highly heritable, so it follows that income will be highly heritable, too." But Bowles and Gintis show that the heritability of income is far larger than the IQ effect can explain. (See Arnold's doubts here).