After the publication of The Son Also Rises, Greg Cochran enjoined Gregory Clark to start swotting up on “Falconer right now”. Interestingly, while looking through some of the references behind Steven Pinker’s (insubstantial) critique of Clark’s “survival of the richest” argument in A Farewell to Alms, I came across this defence of his views by Clark himself. He answers criticisms by Deirdre McCloskey, echoeing those made by Samuel Bowles.
(Here is a 14-page PDF excerpt that deals specifically with the genetic issues.)
Besides demonstrating that McCloskey does not know what “regression to the mean” means, the piece prefigures many of the arguments Clark would later make in A Son Also Rises. He performs a back-of-the-envelope-style calculation of heritability of wealth (while at first staying agnostic about whether the transmission is cultural or economic) in England. Then he speculates whether this could have been genetic. He posits a generic complex trait Z and asks how much selection over how many generations would be required to increase that trait in the population :
With these assumptions the average level of [income] and hence of the trait Z, in the next generation increases by nearly 7% on the base. Thus even one generation in pre-industrial England is enough to change the distribution of the expressed trait significantly. Assuming an average generation length of 33 years there would be 18 generations between 1200 and 1800, plenty of time for significant changes in peoples’ economic aptitudes. If we take the much longer interval between the Neolithic Revolution and the Industrial Revolution we are talking about 8,000 years, 240 generations. That is time enough for quite significant changes in the way people behaved.
As is well known, in the book and the paper “Genetically Capitalist?”, Clark argued that the richest produced more children, and they, by the process of downward mobility, repopulated England. In the defense against McCloskey and Bowles, he illustrates that by showing his heritability exercise would reduce the variance of the trait “economic abilities” :
Another interesting feature of this change is that the variance of economic abilities declines as a result of this selection process. From one generation to the next people become more alike in their underlying economic abilities Z, and that the distribution of y becomes more peaked over time. The evolutionary process would not just raise the mean, it would also reducing the dispersion.
The above is basically the breeders’ equation :