Tax Amnesties as Asset-Laundering Devices

Abstract

This article analyzes the effect of a one-time tax amnesty intended to induce private citizens to voluntarily declare black assets accumulated from past tax evasion. In a dynamic setting, we find that such an amnesty tends to enhance voluntary taxpayer compliance and reduce the size of the underground economy. However, it also generally tends to reduce aggregate net revenues collected by the government in the long run, provided black assets command a higher pretax rate of return than white assets. This is due to a post-amnesty reduction in involuntary revenue collections that outweighs direct amnesty receipts and the increase in voluntary compliance. The desirability of such amnesties thus depends on the importance of unearthing black assets per se, relative to increasing revenue collections. High levels of income tax evasion often result from failure to impose strong penalties for tax offenses. This view obtains support from theoretical analyses of Becker (1968) and Allingham and Sandmo (1972) that predict that tax evasion is unambiguously reduced by higher penalties, thus suggesting that enforcement policies should optimally involve the use of maximal penalties. In practice, penalties for most offenses tend to be nonmaximal. A large body of literature has explored modifications to the Becker-Allingham-Sandmo model to explain why optimal penalties may be nonmaximal. Explanations range from heterogeneity and risk-aversion of citizens and the likelihood of mistaken convictions, to the infeasibility of graduating monitoring effort according to the seriousness of the offense, prosecution costs that rise with the penalty, and the impact of the penalty level on the likelihood of conviction (Andreoni, 1991a; Bolton, 1986;