Yes it really means wealth.... The paper distinguishes wealth from income. Apparently wealth follows the Gibbs distribution but not income, or to quote the interview of Chatterjee :
"Aside from these general models, the scientists also discovered some interesting details within their results. When comparing wealth (i.e. one's net worth) with income, they found that wealth is much more unequally distributed than income (wealth models always have lower Pareto exponents, for any society). Also, while most of the data for the models is based on individuals, data from companies also seemed to follow the same models."
And we are not specifically talking about the "laws" of physics but rather with an ideal thermodynamic model: i.e. a mathematical model that "fits" certain thermodynamic systems, such as the distribution of energy in an ideal gas. One of my questions is, if this paper is true for even a few societies, why should wealth distribution imitate this kind of thermodynamic model at all, ever? I guess this is a bit of an epistemological question. I know Doug, you have written about the trickiness of economic models in "Wall Street" so I was also hoping that you might have some thoughts but also some others, Ravi perhaps or Miles.
The thing is the probabilistic models of thermodynamics do accurately describe certain kinds of "naturalistic" systems such as ideal gases. Why should wealth distribution "imitate" such systems ever... even, if only in certain societies?
Notice that the Gibbs distribution only describes certain particular states of an ideal gas, i.e. when there is "equilibrium", which is also a sort of "ideal" assumption that physicists make in order to study their models. I see a problem here in applying such a notion to human societies because it is possible that "equilibrium" only occurs for limited periods in limited societies. It seems to me that "complex" human societies fall apart much more easily than any "ideal model" can track. So the kind of equilibrium that the model in the study assumes may be very local. This doesn't mean that it is not persistent. _If_ it could be shown that wealth distribution was similar in Egypt 1000 CE and Athens 500 CE and Rome first century BCE as it is in Japan 2000 CE and England 1800 CE, and all of these conformed to the Gibbs distribution, wouldn't that be curious? (I doubt we have enough info to show this though so maybe I am making problems where there is no reason. But my questions are hypothetical as well as specific to this study.)
Jerry Monaco