Agent-based Simulation of Revenue Policy

Abstract

The main research attention is focused on social interaction as a complex system of give-and-take actions with different ways to balance costs and rewards and on the problem of increasing the income inequality. The current investigation analyzes the influence of tax-and-transfer policy on the social structure of society. Different tax regimes were tested by means of computer experiments. The mathematical model of conflict interaction (Koshmanenko, 2014) is the theoretical basis of the agent-based simulation in NetLogo. It is shown that intervention into the process of conflict interactions aimed in support of those who are at the social “bottom” results in the emergence of fluctuations. Even a small value of external replenishment destroys the stable mechanism of internal redistribution and leads to an unpredictable chaotic behaviour. Moreover, the amplitude of these fluctuations is growing rapidly as the scope of assistance increases. The current research tries to give insight into the influence of rent-seeking practice on the social structure of society. Rent-seeking means the practice of obtaining wealth not through economically valuable activity but by extracting it from others (privileges, tax havens, offshores, etc.). The analysis of the results of computer simulation demonstrates qualitative changes in the social structure of society, namely: – the rich class includes only “selected” agents; – economically active agents of the middle class practically disappear while the number of economically passive agents increases; – the number of those who are under the poverty line is growing rapidly. This paper is also aimed at highlighting the potential of agent-based modelling as a research tool for analyzing complex social systems at the edge of chaos and drawing attention to the necessity of filling the gap between traditional social education and computer modelling.