A dynamic model of Tax Evasion

TL;DR

This paper develops a dynamic model of tax evasion incorporating income distribution, social network effects, and policy parameters, validated through simulations, to inform strategies for reducing evasion.

Contribution

It introduces a simplified dynamic model that accounts for income heterogeneity, social topology, and time-dependent audits, providing new insights into tax evasion behavior.

Findings

Equilibrium states depend on income distribution and social network structure.

Higher audit probabilities and fines reduce evasion levels.

Model predictions are validated by extensive Monte Carlo simulations.

Abstract

In this paper we present a simplified model of a proportional taxation system where citizens decide whether to pay taxes or evade them. We initially derive a dynamic equation for the fraction of evaders and then present both its critical points and equilibrium stable states for a well-mixed population and for a fixed audit probability. Our theoretical predictions consider different possible income distributions (homogeneous and heterogeneous), time-dependent audit probability and different possible social topologies, by means of selected complex network configurations. All derived results are validated and confirmed by extended Monte Carlo simulations. Finally, some policy implications aimed to reduce tax evasion are suggested, with regards to the tax and fine rates and the audit probability.