As one who is neither a left-libertarian nor a Friedmanite, I found this exchange somewhat interesting.
I find class analysis far too useful and illuminating to be tempted by Friedman’s proposed solution, but I think he has put his finger in a genuine problem: how exactly are we to identify the ruling and ruled classes? We cannot merely identify them as those who on balance benefit or lose out from the existing system, because it might well be that everyone loses out.
Here’s a first stab: the test of whether one is in the ruling class or the ruled class is not whether one is better or worse off (either subjectively or objectively) than one would be without the system, but rather something like this: where there are two groups A and B, and A occupies a superior socioeconomic position relative to B, and A owes its superior socioeconomic position non-accidentally to the systematic exploitation and/or oppression of B, then A is the ruling and B the ruled class, even if A and B would both be better off than they are now without the exploitation/oppression.
I wanted to capture, first, the idea that what matters is relative rather than absolute position, and second, that the ruling group wins out at the expense of the ruled group.
On the other hand, I’m a bit uncomfortable with the idea that everyone in the ruling group has to be wealthier than everyone in the ruled group; that seems wrong. Obviously “non-accidentally” would need to be filled in too. It surely needs fixing in other ways as well; consider it a first draft, and I want to turn it over to more minds. Suggestions?