An Agent-Based Model of Wealth Distribution

We investigate the agent-based modeling technique in a model of wealth distribution. In the first part we discuss this modern approach to economic modeling in the light of two major methodological approaches in the history of economic analysis, classical political economy and neo-classical economics. In the core part of the paper we present a model which belongs to the large group of essentially neo-classical models that neglect work, production, and productive relations, but rather focuses on distributive interactions in a hunter-gatherer society. We obtain interesting dynamics of inequality in the simulation of wealth distribution. We analyze some causal links between the rules and parameters on the one side and the results on the other side. In this way, we can explain some results in terms of the mechanisms generating them instead of just admiring an "emergent structure." The analysis of relative inequality as measured by the Gini coefficient shows an inverse correlation between the average degree of vision (agent's skills) and wealth inequality expressed by the Gini coefficient. We also explored the effects of inheriting initial wealth and vision. Finally, we do not succeed in simulating the Pareto law, thus failing in replicating an empirical pattern of capitalist distribution of wealth.