Tax Avoidance, Evasion, and Administration**We are grateful to Wojciech Kopczuk for valuable research assistance. Helpful comments on an earlier draft were received from Jim Alm, Alan Auerbach, Len Burman, Brian Erard, Martin Feldstein, Firouz Gahvari, Roger Gordon, Jim Hines, Jonathan Kesselman, Louis Kaplow, Jim Poterba, Agnar Sandmo, Dan Shaviro, Eric Toder, Gideon Yaniv, the participants at the NYU Law School Colloquium on Tax Policy and Public Finance and the handbook conference held at the University of California at Berkeley.

Abstract

Tax avoidance and evasion are pervasive in all countries, and tax structures are undoubtedly skewed by this reality. Standard models of taxation and their conclusions must reflect these realities. This paper first presents theoretical models that integrate avoidance and evasion into the overall decision problem faced by individuals. Early models of this area focused on tax evasion, modeled as a gamble against the enforcement capability of the state. More recently, the literature has examined more general models of the technology of avoidance, with the additional risk bearing caused by tax evasion either being a special case of this technology or one aspect of the cost of changing behavior to reduce tax liability. If the cost of evasion and avoidance depends on other aspects of behavior, the choice of consumption basket and avoidance become intertwined. The paper then relates the behavior predicted by the model to what is known empirically about the extent of evasion and avoidance, and how it responds to tax enforcement policy. The paper then turns to normative analysis, and discusses how avoidance and evasion affect the analysis of vertical and horizontal equity as well as efficiency costs; a taxonomy of efficiency costs is presented. Acknowledging the variety of behavioral responses to taxation changes the answers to traditional subjects of inquiry, such as incidence, optimal progressivity, and the optimal mix between income and consumption taxes. It also raises a whole new set of policy questions, such as the appropriate level of resources to devote to administration and enforcement, and how those resources should be deployed. Because there are a variety of policy instruments that can affect the magnitude and nature of avoidance and evasion response, the elasticity of behavioral response is itself a policy instrument, to be chosen optimally. The paper reviews what is known about these issues, and introduces a general theory of optimal tax systems, in which tax rates and bases are chosen simultaneously with the administrative and enforcement regimes. We argue that the concept of the marginal efficiency cost of funds is a useful way to summarize the normative issues that arise, and expand the concept to include administrative costs, avoidance, and evasion.