Theory and Agent-Based Modeling of Taxpayer Preference and Behavior

The relationship between tax rates and revenue guides governments in making policy decisions regarding tax reforms. The Laffer curve suggests the possibility of an inverse relationship between tax rates and revenue. There are conflicting views and opinions regarding this relationship and the shape of the Laffer curve. Since the behavioral responses of taxpayers to taxation is convoluted, with individuals having their own preferences, it is possible to use agent-based modeling. This paper creates an agent-based model that simulates taxpayer behavior to explore the relationship between tax rates and revenue. Two fundamental responses to taxation: alteration of labor supply, and tax evasion, are simulated in our agent-based model. The revenue collected for different tax rates is plotted and by doing so, the shape of the Laffer curve is verified. We estimate the unemployment rate for varying tax rates and explain the rationale behind this relationship, and further examine the direct and indirect effects of non-labor income taxes on revenue by plotting the revenue collected against labor income tax rates for different tax rates on non-labor income. To estimate the effect of audit rates on revenue, we analyse the percentage of revenue evaded for different audit rates, and finally numerically estimate the satisfaction of taxpayers at varying tax rates.