Taxation and evasion: a dynamic model

In this paper we study tax evasion by means of a unified framework, based on a behavioral approach, where each individual decision with respect to tax compliance is driven by either personal evaluations of the available information, correlated to income and the perception of the quality of the public good, and social influences, derived by the known decision of neighbors. Our model relies on individual utility functions and describes the tax-evasion problem by means of a personal evolutionary scheme, in which each citizen dynamically adapts her behavior as a response to changing economic and social factors. We will show basic economic intuitions on the relevance of penalties, imitation, satisfaction and risk aversion by means of an analytical model and its agent-based companion version, in order to analyse different elements influencing tax evasion and their dynamic effects. In particular, it is shown how the tax-evasion probability changes as a function of the risk-aversion and specific focus is dedicated to the role played by non-monetary elements of utility in inducing high levels of tax compliance also at substantially reduced fines.