Derivation of wealth distributions from biased exchange of money (Q6106922)

The article provides for a relevant contribution to the complexity approach to economics and the econophysics literature, it also provides for a methodological example of the application of the propagation of chaos to economics, being of interest for economics researchers, complexity researchers, and econophysics researchers as well as for people researching in public policies. The authors work with an agent-based model of money exchange in a closed economical system, so that the total wealth is conserved. Three types of exchange interactions between agents are considered: unbiased exchange, poor-biased exchange and rich-biased exchange. In the unbiased exchange model at a random time, selected in accordance with the exponential probability distribution, an agent is chosen at random, with uniform probability distribution, if that agent has at least one dollar it gives that dollar to another agent picked at random also with uniform probability, if that agent's wealth is zero then the agent does nothing. In the poor-biased model, the agent selection probability is no longer uniform, so the agents are randomly selected in proportion to their wealth which means that poor agents have a lower probability of being selected to give a dollar. Finally, in the rich-biased model, the agent selection probability is inversely proportional to the wealth, so that the rich have a lower probability of being selected to give a dollar. The authors simulate the three exchange models showing, through the Gini index, as expected, that the poor-biased model leads to a lower inequality, the unbiased model leads to a higher inequality than the poor-biased model, and the rich-biased model leads to an even higher inequality than the other two models. After discussing the results from the simulations, the authors address the asymptotic dynamics, in this case, the methodology applied is of special interest for researchers in the complexity sciences, since it deals with generalizable methodologies applicable to the research on the asymptotic dynamics of agent-based systems addressable in terms of statistical mechanics' applications. In this case, to research the asymptotic dynamics, instead of taking the large time limit and then the population limit, the authors take the population limit and then the time limit applying the propagation of chaos method. The main point in the propagation of chaos is that by taking the large population limit one is able to go from a multiagent (large) stochastic system to a deterministic system of differential equations, a point reviewed by the authors in regards to this method. In the case of the article's model, the authors apply this approach taking first the population limit with the number of agents tending to infinity and then analyzing the deterministic description in terms of a discrete Fokker-Planck formulation then, using entropy methods, the convergence to an equilibrium distribution is researched. In Section 1, the authors provide for the main structure of the model, methods and main findings. In Section 2, a formal and general review of the propagation of chaos is provided. In Section 3, the unbiased exchange model is addressed in terms of its asymptotic behavior, recovering major previous findings of the literature on the subject. In Section 4, the poor-biased model is researched and found to have a unique equilibrium given by a Poisson distribution, the convergence to the equilibrium is studied both formally and numerically. The rich-biased model is researched in Section 5, in this case, the authors uncover, in the numerical simulations, not only a dynamics of rising inequality, but also a travelling wave dynamics. In this case, the dynamics is such that the poor get poorer and the rich get richer, however, the proportion of poor increases with time while the dynamics associated with the rich follows the pattern of a dispersive travelling wave. The authors address this dynamics in some detail and the findings are of particular importance to economics since they demonstrate the existence of an unsustainability in the rich-biased dynamics in comparison to the other two models, namely, this dynamics as shown and stressed by the authors does not converge strongly to an equilibrium and the travelling wave leads also to a dispersion of poverty with the rich getting richer and in smaller proportion of the population leading to a vanishing wealth, this wave dynamics also leads the Gini index to converge to its maximum value. From an economics and public policy perspective, this result is of key importance. From a complexity research standpoint the uncovering of the travelling wave dynamics in this agent-based system of exchange is of particular importance and needs further analytical study, a point that the authors stress in the fifth and last section of the article. Overall, the article provides for an important contribution to the field of econophysics, both methodologically and in terms of the results reached by the authors for the wealth distribution problem, with potential impact both for economics and public policy research.