Hm. I'm having a hard time wrapping my mind around this. Please elaborate, if you will?

Gladly. (EDITED: I originally used 'wealth' and 'power' instead of 'wealth' and 'coercion'. It was rightly pointed out that 'wealth' is just a form of 'power'. So rather than "separating wealth from power", what I'm really talking about is "separating our ideas about wealth from our ideas about coercion".)

On one axis, we have poverty/wealth, which is a measure of "how much stuff do I have?" - how much food, water, air, shelter, etc. - up to and including how much control I have over my physical environment. Near one end of the spectrum, we can imagine skeletal children waiting for vultures to eat them. Near the other end of the spectrum, we can imagine eloi living within utility fogs, waiting to condense anything they dream of out of thin air.

On another axis, we have the social ladder / status games that humans play so well. Near one end of that spectrum, we can imagine abject slaves, whose right to live is up to the whim of others; near the other end of that spectrum, we can imagine despotic conquerors, whose whim controls the lives of millions.

The poverty/wealth spectrum is NOT zero-sum, because even if we collapse a society's span on that spectrum down to a single point, WHERE that point falls matters - everyone being equally poor is materially different from everyone being equally rich.

On the other hand, the social coercion spectrum is fundamentally zero-sum - you can't remove B's inferiority to A, without removing A's superiority to B.

The problem is, most people don't seem to REALLY separate those two spectra - when we talk about "making people more equal", we usually talk about wealth redistribution, not coercion redistribution. Paul Graham discusses this at length.

The thing is, coercion influences wealth, and vice-versa. But the fact that they influence each other doesn't make them the same thing - it just means that the phase diagram covers a subset of the phase space, instead of the whole thing.

One thing you notice when you look at history, is how various technologies (which are physical objects, and therefore on the wealth spectrum) influence the course of society and governance (which are social objects, and therefore on the coercion spectrum). For example, look at how readily "freer" democratic governments break out when new military technology makes it easier to mount rebellions, and how readily authoritarian regimes break out when new military technology makes it easier to consolidate coercion. And look at how readily technological revolutions break out when a culture allows free exploration of possibilities, and how readily they die when a culture imposes direct control over what its entrepreneurs and researchers can study and make.

A testable claim that the Progressive movement might make, if these two continua are recognized as separate, is that minimizing the coercion disparity between individuals will tend to maximize their aggregate wealth, while maximizing the coercion disparity between individuals will tend to lower that aggregate wealth. The American version of Libertarianism seems to make the opposite claim, that allowing power disparities to naturally follow the individuals' innate abilities to seize and hold power will allow those individuals who are better at multiplying wealth to do so, while redistributing their wealth to those with less innate ability to seize and hold power is inefficient.

Even better, though, separating out those two concepts allows certain testable claims to be made more clearly - things like "we can redistribute coercive power without redistributing wealth, and regulation is one way to do that", and "all attempts to redistribute or nullify coercive power must use coercive power to do so, and will therefore favor those who currently have coercive power".

EDIT: Upon further reflection, I think it's not entirely true that people can't distinguish between these two concepts - plenty of people can and do distinguish between them, but then when they come to making policy decisions, they conflate them again. This looks like a cognitive bias at work, but I'm not sure which one - something that causes people to default to zero-sum thinking even when they know a situation has positive-sum solutions?