[dwallin]

A great book, but I would love to see additional studies really breaking it down. My gut feeling is that this effect curves. At a certain size of selectorate you switch from direct influence, with linear costs, to indirect methods of influence, with much better scaling properties (eg. Uber and California Proposition 22).

While it becomes expensive to outright buy people, it also becomes prohibitively expensive to educate, and an ability to outspend with emotional appeals becomes more important than having a cogent argument.