Agent-based Modeling of Tax Evasion

Preface

Tax evasion or tax noncompliance is an age-old behavior of humans. It is probably as old as taxation itself. Taxation is an essential tool of any institutionalized community, a means by which resources are collected for the common causes. However, free-riding also appears to be an age-old, ever-existing phenomenon, leading to actors seeking safe ways to avoid the burdens of taxation. Various rulers and governments developed various taxes and tax systems, adding several layers of interpretations for avoiding paying taxes. Subjects of unjust rulers may feel disassociated with the projects that are financed by their taxes and may also feel that they do not possess the power to influence the decisions on the allocation of the taxes collected. This leads to a very complex net of possible reasons, causes, and effects for tax noncompliance that culminate in the repeated decisions by millions of individuals. Such decisions lead to the numbers individuals put in their tax reporting forms.

Agent-based modeling, on the other hand, is a novel methodology to study complex social systems, made possible by the recent advances in computer technology. Its main tenet is to model the individual decision maker, with all its idiosyncrasies and interactions, with specific goals and concerns. The multi-level, networked interactions of these individually modeled decisions are often too complex to follow with the analytical means provided by traditional mathematics. Therefore, the consequences ...