There Is No Spoon : Reconsidering the Tax Compliance Puzzle

For over 40 years theorists have sought the effects of tax audits on voluntary compliance rates by studying individual taxpayer motivations. Yet no single theory has produced a taxpayer incentive model that both comports with experience and explains the effects of audits on compliance. This quandary is often termed the “tax compliance puzzle.” Consequently, some theorists have called for more capacious models that make room for the panoply of individual compliance motivations. This paper proposes that a more complex model is unnecessary. To the contrary, complex compliance and enforcement data can result from extremely simple behavioral rules of individual taxpayers and government examiners interacting over time. This paper describes an agent-based computational model that uses a single, simple rule of action for each taxpayer and examiner. The model produces three interesting effects supporting the conclusion that there may be no tax compliance puzzle to solve. First, the results comport with known U.S. compliance and audit rates. Second, the results suggest that while audit probability influences individual compliance decisions, it has negligible effects on system-level compliance patterns. Third, the results support the theory that the perceived strength of the tax authority correlates directly — but nonlinearly — with voluntary compliance rates. The model is not complete enough to determine conclusively that this last effect is due to perceived strength of the tax authority alone and might be due instead to factors such as social norms and other behaviorist theories